diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,17340 @@
+{
+  "best_metric": 0.48,
+  "best_model_checkpoint": "5c_4/checkpoint-9360",
+  "epoch": 99.01,
+  "eval_steps": 500,
+  "global_step": 23400,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00042735042735042735,
+      "grad_norm": 6.5477800369262695,
+      "learning_rate": 4.273504273504274e-08,
+      "loss": 1.341,
+      "step": 10
+    },
+    {
+      "epoch": 0.0008547008547008547,
+      "grad_norm": 7.9439568519592285,
+      "learning_rate": 8.547008547008549e-08,
+      "loss": 1.3047,
+      "step": 20
+    },
+    {
+      "epoch": 0.001282051282051282,
+      "grad_norm": 6.425937175750732,
+      "learning_rate": 1.282051282051282e-07,
+      "loss": 1.3641,
+      "step": 30
+    },
+    {
+      "epoch": 0.0017094017094017094,
+      "grad_norm": 6.959607124328613,
+      "learning_rate": 1.7094017094017097e-07,
+      "loss": 1.3257,
+      "step": 40
+    },
+    {
+      "epoch": 0.002136752136752137,
+      "grad_norm": 6.3154730796813965,
+      "learning_rate": 2.136752136752137e-07,
+      "loss": 1.3008,
+      "step": 50
+    },
+    {
+      "epoch": 0.002564102564102564,
+      "grad_norm": 6.919759750366211,
+      "learning_rate": 2.564102564102564e-07,
+      "loss": 1.3322,
+      "step": 60
+    },
+    {
+      "epoch": 0.0029914529914529917,
+      "grad_norm": 6.870538711547852,
+      "learning_rate": 2.991452991452992e-07,
+      "loss": 1.313,
+      "step": 70
+    },
+    {
+      "epoch": 0.003418803418803419,
+      "grad_norm": 6.823713302612305,
+      "learning_rate": 3.4188034188034194e-07,
+      "loss": 1.3179,
+      "step": 80
+    },
+    {
+      "epoch": 0.0038461538461538464,
+      "grad_norm": 7.863243579864502,
+      "learning_rate": 3.846153846153847e-07,
+      "loss": 1.2466,
+      "step": 90
+    },
+    {
+      "epoch": 0.004273504273504274,
+      "grad_norm": 7.5396881103515625,
+      "learning_rate": 4.273504273504274e-07,
+      "loss": 1.2098,
+      "step": 100
+    },
+    {
+      "epoch": 0.004700854700854701,
+      "grad_norm": 8.087361335754395,
+      "learning_rate": 4.700854700854701e-07,
+      "loss": 1.2309,
+      "step": 110
+    },
+    {
+      "epoch": 0.005128205128205128,
+      "grad_norm": 8.429758071899414,
+      "learning_rate": 5.128205128205128e-07,
+      "loss": 1.1752,
+      "step": 120
+    },
+    {
+      "epoch": 0.005555555555555556,
+      "grad_norm": 9.436071395874023,
+      "learning_rate": 5.555555555555555e-07,
+      "loss": 1.1474,
+      "step": 130
+    },
+    {
+      "epoch": 0.005982905982905983,
+      "grad_norm": 80.47299194335938,
+      "learning_rate": 5.982905982905984e-07,
+      "loss": 1.1764,
+      "step": 140
+    },
+    {
+      "epoch": 0.00641025641025641,
+      "grad_norm": 16.705190658569336,
+      "learning_rate": 6.41025641025641e-07,
+      "loss": 1.1038,
+      "step": 150
+    },
+    {
+      "epoch": 0.006837606837606838,
+      "grad_norm": 22.808284759521484,
+      "learning_rate": 6.837606837606839e-07,
+      "loss": 1.1758,
+      "step": 160
+    },
+    {
+      "epoch": 0.007264957264957265,
+      "grad_norm": 23.328384399414062,
+      "learning_rate": 7.264957264957266e-07,
+      "loss": 1.2182,
+      "step": 170
+    },
+    {
+      "epoch": 0.007692307692307693,
+      "grad_norm": 16.98978614807129,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 0.8752,
+      "step": 180
+    },
+    {
+      "epoch": 0.00811965811965812,
+      "grad_norm": 197.35919189453125,
+      "learning_rate": 8.11965811965812e-07,
+      "loss": 1.122,
+      "step": 190
+    },
+    {
+      "epoch": 0.008547008547008548,
+      "grad_norm": 32.2603759765625,
+      "learning_rate": 8.547008547008548e-07,
+      "loss": 1.1263,
+      "step": 200
+    },
+    {
+      "epoch": 0.008974358974358974,
+      "grad_norm": 15.653111457824707,
+      "learning_rate": 8.974358974358975e-07,
+      "loss": 1.3598,
+      "step": 210
+    },
+    {
+      "epoch": 0.009401709401709401,
+      "grad_norm": 13.729981422424316,
+      "learning_rate": 9.401709401709402e-07,
+      "loss": 0.6189,
+      "step": 220
+    },
+    {
+      "epoch": 0.009829059829059829,
+      "grad_norm": 30.82917594909668,
+      "learning_rate": 9.829059829059829e-07,
+      "loss": 0.9641,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 1.4837762117385864,
+      "eval_runtime": 34.3992,
+      "eval_samples_per_second": 0.727,
+      "eval_steps_per_second": 0.727,
+      "step": 234
+    },
+    {
+      "epoch": 1.0002564102564102,
+      "grad_norm": 13.610635757446289,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 0.9614,
+      "step": 240
+    },
+    {
+      "epoch": 1.0006837606837606,
+      "grad_norm": 51.50576400756836,
+      "learning_rate": 1.0683760683760685e-06,
+      "loss": 0.9305,
+      "step": 250
+    },
+    {
+      "epoch": 1.001111111111111,
+      "grad_norm": 101.85962677001953,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 0.923,
+      "step": 260
+    },
+    {
+      "epoch": 1.0015384615384615,
+      "grad_norm": 15.406578063964844,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 1.558,
+      "step": 270
+    },
+    {
+      "epoch": 1.001965811965812,
+      "grad_norm": 49.64127731323242,
+      "learning_rate": 1.1965811965811968e-06,
+      "loss": 0.749,
+      "step": 280
+    },
+    {
+      "epoch": 1.0023931623931623,
+      "grad_norm": 23.893503189086914,
+      "learning_rate": 1.2393162393162394e-06,
+      "loss": 0.7217,
+      "step": 290
+    },
+    {
+      "epoch": 1.0028205128205128,
+      "grad_norm": 12.812580108642578,
+      "learning_rate": 1.282051282051282e-06,
+      "loss": 1.3277,
+      "step": 300
+    },
+    {
+      "epoch": 1.0032478632478632,
+      "grad_norm": 8.36343002319336,
+      "learning_rate": 1.324786324786325e-06,
+      "loss": 0.4727,
+      "step": 310
+    },
+    {
+      "epoch": 1.0036752136752136,
+      "grad_norm": 3.717958450317383,
+      "learning_rate": 1.3675213675213678e-06,
+      "loss": 1.0901,
+      "step": 320
+    },
+    {
+      "epoch": 1.004102564102564,
+      "grad_norm": 120.84101867675781,
+      "learning_rate": 1.4102564102564104e-06,
+      "loss": 1.2407,
+      "step": 330
+    },
+    {
+      "epoch": 1.0045299145299145,
+      "grad_norm": 76.76756286621094,
+      "learning_rate": 1.4529914529914531e-06,
+      "loss": 1.2918,
+      "step": 340
+    },
+    {
+      "epoch": 1.004957264957265,
+      "grad_norm": 66.22538757324219,
+      "learning_rate": 1.4957264957264957e-06,
+      "loss": 2.1196,
+      "step": 350
+    },
+    {
+      "epoch": 1.0053846153846153,
+      "grad_norm": 49.87553024291992,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 0.9441,
+      "step": 360
+    },
+    {
+      "epoch": 1.0058119658119657,
+      "grad_norm": 64.26871490478516,
+      "learning_rate": 1.5811965811965813e-06,
+      "loss": 1.3502,
+      "step": 370
+    },
+    {
+      "epoch": 1.0062393162393162,
+      "grad_norm": 1.687718391418457,
+      "learning_rate": 1.623931623931624e-06,
+      "loss": 2.3065,
+      "step": 380
+    },
+    {
+      "epoch": 1.0066666666666666,
+      "grad_norm": 1.3850126266479492,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.9136,
+      "step": 390
+    },
+    {
+      "epoch": 1.007094017094017,
+      "grad_norm": 1.278273344039917,
+      "learning_rate": 1.7094017094017097e-06,
+      "loss": 1.5894,
+      "step": 400
+    },
+    {
+      "epoch": 1.0075213675213675,
+      "grad_norm": 63.78499221801758,
+      "learning_rate": 1.7521367521367522e-06,
+      "loss": 1.1296,
+      "step": 410
+    },
+    {
+      "epoch": 1.0079487179487179,
+      "grad_norm": 0.9789712429046631,
+      "learning_rate": 1.794871794871795e-06,
+      "loss": 1.6205,
+      "step": 420
+    },
+    {
+      "epoch": 1.0083760683760683,
+      "grad_norm": 0.5999449491500854,
+      "learning_rate": 1.8376068376068378e-06,
+      "loss": 1.7319,
+      "step": 430
+    },
+    {
+      "epoch": 1.0088034188034187,
+      "grad_norm": 0.31746140122413635,
+      "learning_rate": 1.8803418803418804e-06,
+      "loss": 0.643,
+      "step": 440
+    },
+    {
+      "epoch": 1.0092307692307692,
+      "grad_norm": 76.34538269042969,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 2.4496,
+      "step": 450
+    },
+    {
+      "epoch": 1.0096581196581196,
+      "grad_norm": 5.805349349975586,
+      "learning_rate": 1.9658119658119658e-06,
+      "loss": 1.5439,
+      "step": 460
+    },
+    {
+      "epoch": 1.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 3.7125139236450195,
+      "eval_runtime": 33.4952,
+      "eval_samples_per_second": 0.746,
+      "eval_steps_per_second": 0.746,
+      "step": 468
+    },
+    {
+      "epoch": 2.00008547008547,
+      "grad_norm": 72.71012878417969,
+      "learning_rate": 2.008547008547009e-06,
+      "loss": 1.7175,
+      "step": 470
+    },
+    {
+      "epoch": 2.0005128205128204,
+      "grad_norm": 0.6215125918388367,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 2.393,
+      "step": 480
+    },
+    {
+      "epoch": 2.000940170940171,
+      "grad_norm": 1.5661790370941162,
+      "learning_rate": 2.094017094017094e-06,
+      "loss": 2.3866,
+      "step": 490
+    },
+    {
+      "epoch": 2.0013675213675213,
+      "grad_norm": 0.6268726587295532,
+      "learning_rate": 2.136752136752137e-06,
+      "loss": 2.4783,
+      "step": 500
+    },
+    {
+      "epoch": 2.0017948717948717,
+      "grad_norm": 95.59490203857422,
+      "learning_rate": 2.1794871794871797e-06,
+      "loss": 2.3849,
+      "step": 510
+    },
+    {
+      "epoch": 2.002222222222222,
+      "grad_norm": 0.4951792061328888,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 1.2576,
+      "step": 520
+    },
+    {
+      "epoch": 2.0026495726495726,
+      "grad_norm": 0.2503603398799896,
+      "learning_rate": 2.2649572649572653e-06,
+      "loss": 1.2306,
+      "step": 530
+    },
+    {
+      "epoch": 2.003076923076923,
+      "grad_norm": 52.469234466552734,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 1.8782,
+      "step": 540
+    },
+    {
+      "epoch": 2.0035042735042734,
+      "grad_norm": 0.20398114621639252,
+      "learning_rate": 2.3504273504273504e-06,
+      "loss": 0.5924,
+      "step": 550
+    },
+    {
+      "epoch": 2.003931623931624,
+      "grad_norm": 71.42668914794922,
+      "learning_rate": 2.3931623931623937e-06,
+      "loss": 3.2851,
+      "step": 560
+    },
+    {
+      "epoch": 2.0043589743589743,
+      "grad_norm": 56.71471405029297,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 1.2107,
+      "step": 570
+    },
+    {
+      "epoch": 2.0047863247863247,
+      "grad_norm": 0.25754937529563904,
+      "learning_rate": 2.478632478632479e-06,
+      "loss": 1.874,
+      "step": 580
+    },
+    {
+      "epoch": 2.005213675213675,
+      "grad_norm": 1.0091850757598877,
+      "learning_rate": 2.5213675213675216e-06,
+      "loss": 2.3834,
+      "step": 590
+    },
+    {
+      "epoch": 2.0056410256410255,
+      "grad_norm": 0.3749796450138092,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 1.1427,
+      "step": 600
+    },
+    {
+      "epoch": 2.006068376068376,
+      "grad_norm": 0.9544910788536072,
+      "learning_rate": 2.606837606837607e-06,
+      "loss": 2.4185,
+      "step": 610
+    },
+    {
+      "epoch": 2.0064957264957264,
+      "grad_norm": 0.21911416947841644,
+      "learning_rate": 2.64957264957265e-06,
+      "loss": 1.2006,
+      "step": 620
+    },
+    {
+      "epoch": 2.006923076923077,
+      "grad_norm": 0.1923692524433136,
+      "learning_rate": 2.6923076923076923e-06,
+      "loss": 1.2047,
+      "step": 630
+    },
+    {
+      "epoch": 2.0073504273504272,
+      "grad_norm": 0.2782430052757263,
+      "learning_rate": 2.7350427350427355e-06,
+      "loss": 1.7899,
+      "step": 640
+    },
+    {
+      "epoch": 2.0077777777777777,
+      "grad_norm": 0.17398680746555328,
+      "learning_rate": 2.7777777777777783e-06,
+      "loss": 1.0863,
+      "step": 650
+    },
+    {
+      "epoch": 2.008205128205128,
+      "grad_norm": 121.83155059814453,
+      "learning_rate": 2.8205128205128207e-06,
+      "loss": 1.805,
+      "step": 660
+    },
+    {
+      "epoch": 2.0086324786324785,
+      "grad_norm": 6.328676700592041,
+      "learning_rate": 2.8632478632478635e-06,
+      "loss": 1.6228,
+      "step": 670
+    },
+    {
+      "epoch": 2.009059829059829,
+      "grad_norm": 74.6235580444336,
+      "learning_rate": 2.9059829059829063e-06,
+      "loss": 3.3584,
+      "step": 680
+    },
+    {
+      "epoch": 2.0094871794871794,
+      "grad_norm": 0.6486341953277588,
+      "learning_rate": 2.948717948717949e-06,
+      "loss": 0.8365,
+      "step": 690
+    },
+    {
+      "epoch": 2.00991452991453,
+      "grad_norm": 0.20082318782806396,
+      "learning_rate": 2.9914529914529914e-06,
+      "loss": 1.2944,
+      "step": 700
+    },
+    {
+      "epoch": 2.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 3.6749236583709717,
+      "eval_runtime": 33.4285,
+      "eval_samples_per_second": 0.748,
+      "eval_steps_per_second": 0.748,
+      "step": 702
+    },
+    {
+      "epoch": 3.0003418803418804,
+      "grad_norm": 49.130332946777344,
+      "learning_rate": 3.0341880341880342e-06,
+      "loss": 3.0762,
+      "step": 710
+    },
+    {
+      "epoch": 3.000769230769231,
+      "grad_norm": 10.840944290161133,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 3.0125,
+      "step": 720
+    },
+    {
+      "epoch": 3.0011965811965813,
+      "grad_norm": 0.9621607661247253,
+      "learning_rate": 3.11965811965812e-06,
+      "loss": 1.0516,
+      "step": 730
+    },
+    {
+      "epoch": 3.0016239316239317,
+      "grad_norm": 0.4117166996002197,
+      "learning_rate": 3.1623931623931626e-06,
+      "loss": 0.5509,
+      "step": 740
+    },
+    {
+      "epoch": 3.002051282051282,
+      "grad_norm": 0.34464403986930847,
+      "learning_rate": 3.205128205128206e-06,
+      "loss": 0.5911,
+      "step": 750
+    },
+    {
+      "epoch": 3.0024786324786326,
+      "grad_norm": 0.21714632213115692,
+      "learning_rate": 3.247863247863248e-06,
+      "loss": 1.2975,
+      "step": 760
+    },
+    {
+      "epoch": 3.002905982905983,
+      "grad_norm": 0.6396281719207764,
+      "learning_rate": 3.290598290598291e-06,
+      "loss": 3.169,
+      "step": 770
+    },
+    {
+      "epoch": 3.0033333333333334,
+      "grad_norm": 0.28683122992515564,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.8898,
+      "step": 780
+    },
+    {
+      "epoch": 3.003760683760684,
+      "grad_norm": 0.158400759100914,
+      "learning_rate": 3.3760683760683765e-06,
+      "loss": 0.6267,
+      "step": 790
+    },
+    {
+      "epoch": 3.0041880341880343,
+      "grad_norm": 49.82303237915039,
+      "learning_rate": 3.4188034188034193e-06,
+      "loss": 0.6848,
+      "step": 800
+    },
+    {
+      "epoch": 3.0046153846153847,
+      "grad_norm": 0.5617250800132751,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 3.2734,
+      "step": 810
+    },
+    {
+      "epoch": 3.005042735042735,
+      "grad_norm": 104.6950912475586,
+      "learning_rate": 3.5042735042735045e-06,
+      "loss": 2.4173,
+      "step": 820
+    },
+    {
+      "epoch": 3.0054700854700855,
+      "grad_norm": 3.587768077850342,
+      "learning_rate": 3.5470085470085473e-06,
+      "loss": 0.7486,
+      "step": 830
+    },
+    {
+      "epoch": 3.005897435897436,
+      "grad_norm": 1.223972201347351,
+      "learning_rate": 3.58974358974359e-06,
+      "loss": 2.2701,
+      "step": 840
+    },
+    {
+      "epoch": 3.0063247863247864,
+      "grad_norm": 84.18639373779297,
+      "learning_rate": 3.632478632478633e-06,
+      "loss": 2.5556,
+      "step": 850
+    },
+    {
+      "epoch": 3.006752136752137,
+      "grad_norm": 0.33438971638679504,
+      "learning_rate": 3.6752136752136756e-06,
+      "loss": 1.1486,
+      "step": 860
+    },
+    {
+      "epoch": 3.0071794871794872,
+      "grad_norm": 0.4742439091205597,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 1.8177,
+      "step": 870
+    },
+    {
+      "epoch": 3.0076068376068377,
+      "grad_norm": 0.20394758880138397,
+      "learning_rate": 3.760683760683761e-06,
+      "loss": 1.2203,
+      "step": 880
+    },
+    {
+      "epoch": 3.008034188034188,
+      "grad_norm": 51.59945297241211,
+      "learning_rate": 3.8034188034188036e-06,
+      "loss": 1.09,
+      "step": 890
+    },
+    {
+      "epoch": 3.0084615384615385,
+      "grad_norm": 0.2848564386367798,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 1.3275,
+      "step": 900
+    },
+    {
+      "epoch": 3.008888888888889,
+      "grad_norm": 0.22904685139656067,
+      "learning_rate": 3.88888888888889e-06,
+      "loss": 3.0584,
+      "step": 910
+    },
+    {
+      "epoch": 3.0093162393162394,
+      "grad_norm": 33.74935531616211,
+      "learning_rate": 3.9316239316239315e-06,
+      "loss": 1.7752,
+      "step": 920
+    },
+    {
+      "epoch": 3.00974358974359,
+      "grad_norm": 0.3561188280582428,
+      "learning_rate": 3.974358974358974e-06,
+      "loss": 0.9419,
+      "step": 930
+    },
+    {
+      "epoch": 3.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 3.042177200317383,
+      "eval_runtime": 33.5077,
+      "eval_samples_per_second": 0.746,
+      "eval_steps_per_second": 0.746,
+      "step": 936
+    },
+    {
+      "epoch": 4.00017094017094,
+      "grad_norm": 61.73008728027344,
+      "learning_rate": 4.017094017094018e-06,
+      "loss": 3.1491,
+      "step": 940
+    },
+    {
+      "epoch": 4.00059829059829,
+      "grad_norm": 32.0042839050293,
+      "learning_rate": 4.05982905982906e-06,
+      "loss": 1.1443,
+      "step": 950
+    },
+    {
+      "epoch": 4.001025641025641,
+      "grad_norm": 0.49480342864990234,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 1.1544,
+      "step": 960
+    },
+    {
+      "epoch": 4.001452991452991,
+      "grad_norm": 0.1361558586359024,
+      "learning_rate": 4.145299145299146e-06,
+      "loss": 0.574,
+      "step": 970
+    },
+    {
+      "epoch": 4.001880341880342,
+      "grad_norm": 0.10107045620679855,
+      "learning_rate": 4.188034188034188e-06,
+      "loss": 2.0072,
+      "step": 980
+    },
+    {
+      "epoch": 4.002307692307692,
+      "grad_norm": 36.34455871582031,
+      "learning_rate": 4.230769230769231e-06,
+      "loss": 1.849,
+      "step": 990
+    },
+    {
+      "epoch": 4.0027350427350425,
+      "grad_norm": 179.90792846679688,
+      "learning_rate": 4.273504273504274e-06,
+      "loss": 1.7365,
+      "step": 1000
+    },
+    {
+      "epoch": 4.003162393162393,
+      "grad_norm": 38.759735107421875,
+      "learning_rate": 4.316239316239317e-06,
+      "loss": 2.1145,
+      "step": 1010
+    },
+    {
+      "epoch": 4.003589743589743,
+      "grad_norm": 0.1710689514875412,
+      "learning_rate": 4.358974358974359e-06,
+      "loss": 0.0095,
+      "step": 1020
+    },
+    {
+      "epoch": 4.004017094017094,
+      "grad_norm": 0.4290190041065216,
+      "learning_rate": 4.401709401709402e-06,
+      "loss": 1.7793,
+      "step": 1030
+    },
+    {
+      "epoch": 4.004444444444444,
+      "grad_norm": 32.45518493652344,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 1.7186,
+      "step": 1040
+    },
+    {
+      "epoch": 4.004871794871795,
+      "grad_norm": 36.26081466674805,
+      "learning_rate": 4.487179487179488e-06,
+      "loss": 1.7024,
+      "step": 1050
+    },
+    {
+      "epoch": 4.005299145299145,
+      "grad_norm": 0.3600796163082123,
+      "learning_rate": 4.5299145299145306e-06,
+      "loss": 0.5479,
+      "step": 1060
+    },
+    {
+      "epoch": 4.0057264957264955,
+      "grad_norm": 0.39198485016822815,
+      "learning_rate": 4.5726495726495725e-06,
+      "loss": 1.6426,
+      "step": 1070
+    },
+    {
+      "epoch": 4.006153846153846,
+      "grad_norm": 0.25071772933006287,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 1.0627,
+      "step": 1080
+    },
+    {
+      "epoch": 4.006581196581196,
+      "grad_norm": 35.45331573486328,
+      "learning_rate": 4.658119658119659e-06,
+      "loss": 2.8291,
+      "step": 1090
+    },
+    {
+      "epoch": 4.007008547008547,
+      "grad_norm": 0.4302135407924652,
+      "learning_rate": 4.700854700854701e-06,
+      "loss": 1.1957,
+      "step": 1100
+    },
+    {
+      "epoch": 4.007435897435897,
+      "grad_norm": 35.813323974609375,
+      "learning_rate": 4.743589743589744e-06,
+      "loss": 2.2429,
+      "step": 1110
+    },
+    {
+      "epoch": 4.007863247863248,
+      "grad_norm": 35.34008026123047,
+      "learning_rate": 4.786324786324787e-06,
+      "loss": 1.077,
+      "step": 1120
+    },
+    {
+      "epoch": 4.008290598290598,
+      "grad_norm": 1.468093991279602,
+      "learning_rate": 4.829059829059829e-06,
+      "loss": 2.7016,
+      "step": 1130
+    },
+    {
+      "epoch": 4.0087179487179485,
+      "grad_norm": 29.234296798706055,
+      "learning_rate": 4.871794871794872e-06,
+      "loss": 2.1301,
+      "step": 1140
+    },
+    {
+      "epoch": 4.009145299145299,
+      "grad_norm": 31.05660629272461,
+      "learning_rate": 4.914529914529915e-06,
+      "loss": 2.0786,
+      "step": 1150
+    },
+    {
+      "epoch": 4.009572649572649,
+      "grad_norm": 0.4601377546787262,
+      "learning_rate": 4.957264957264958e-06,
+      "loss": 2.0616,
+      "step": 1160
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 1.4168052673339844,
+      "learning_rate": 5e-06,
+      "loss": 2.4333,
+      "step": 1170
+    },
+    {
+      "epoch": 4.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 2.6802823543548584,
+      "eval_runtime": 33.4138,
+      "eval_samples_per_second": 0.748,
+      "eval_steps_per_second": 0.748,
+      "step": 1170
+    },
+    {
+      "epoch": 5.00042735042735,
+      "grad_norm": 0.4595077633857727,
+      "learning_rate": 5.042735042735043e-06,
+      "loss": 0.8973,
+      "step": 1180
+    },
+    {
+      "epoch": 5.000854700854701,
+      "grad_norm": 32.307037353515625,
+      "learning_rate": 5.085470085470086e-06,
+      "loss": 1.9978,
+      "step": 1190
+    },
+    {
+      "epoch": 5.001282051282051,
+      "grad_norm": 33.618709564208984,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 3.0544,
+      "step": 1200
+    },
+    {
+      "epoch": 5.001709401709402,
+      "grad_norm": 31.92491340637207,
+      "learning_rate": 5.1709401709401716e-06,
+      "loss": 1.8459,
+      "step": 1210
+    },
+    {
+      "epoch": 5.002136752136752,
+      "grad_norm": 0.769111156463623,
+      "learning_rate": 5.213675213675214e-06,
+      "loss": 1.928,
+      "step": 1220
+    },
+    {
+      "epoch": 5.0025641025641026,
+      "grad_norm": 30.650548934936523,
+      "learning_rate": 5.256410256410257e-06,
+      "loss": 1.1278,
+      "step": 1230
+    },
+    {
+      "epoch": 5.002991452991453,
+      "grad_norm": 31.011295318603516,
+      "learning_rate": 5.2991452991453e-06,
+      "loss": 1.8213,
+      "step": 1240
+    },
+    {
+      "epoch": 5.003418803418803,
+      "grad_norm": 0.40149080753326416,
+      "learning_rate": 5.341880341880342e-06,
+      "loss": 2.31,
+      "step": 1250
+    },
+    {
+      "epoch": 5.003846153846154,
+      "grad_norm": 0.4121752977371216,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 1.051,
+      "step": 1260
+    },
+    {
+      "epoch": 5.004273504273504,
+      "grad_norm": 0.29824018478393555,
+      "learning_rate": 5.4273504273504275e-06,
+      "loss": 2.0995,
+      "step": 1270
+    },
+    {
+      "epoch": 5.004700854700855,
+      "grad_norm": 0.4708477258682251,
+      "learning_rate": 5.470085470085471e-06,
+      "loss": 0.4382,
+      "step": 1280
+    },
+    {
+      "epoch": 5.005128205128205,
+      "grad_norm": 0.25618794560432434,
+      "learning_rate": 5.512820512820514e-06,
+      "loss": 2.2577,
+      "step": 1290
+    },
+    {
+      "epoch": 5.0055555555555555,
+      "grad_norm": 0.17655441164970398,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 0.6334,
+      "step": 1300
+    },
+    {
+      "epoch": 5.005982905982906,
+      "grad_norm": 29.26540184020996,
+      "learning_rate": 5.598290598290599e-06,
+      "loss": 1.6359,
+      "step": 1310
+    },
+    {
+      "epoch": 5.006410256410256,
+      "grad_norm": 0.33771806955337524,
+      "learning_rate": 5.641025641025641e-06,
+      "loss": 2.2658,
+      "step": 1320
+    },
+    {
+      "epoch": 5.006837606837607,
+      "grad_norm": 0.19950927793979645,
+      "learning_rate": 5.683760683760684e-06,
+      "loss": 0.6106,
+      "step": 1330
+    },
+    {
+      "epoch": 5.007264957264957,
+      "grad_norm": 32.63151931762695,
+      "learning_rate": 5.726495726495727e-06,
+      "loss": 1.9519,
+      "step": 1340
+    },
+    {
+      "epoch": 5.007692307692308,
+      "grad_norm": 0.3686060309410095,
+      "learning_rate": 5.769230769230769e-06,
+      "loss": 1.2481,
+      "step": 1350
+    },
+    {
+      "epoch": 5.008119658119658,
+      "grad_norm": 0.1446453183889389,
+      "learning_rate": 5.8119658119658126e-06,
+      "loss": 0.5926,
+      "step": 1360
+    },
+    {
+      "epoch": 5.0085470085470085,
+      "grad_norm": 32.86191177368164,
+      "learning_rate": 5.854700854700855e-06,
+      "loss": 2.247,
+      "step": 1370
+    },
+    {
+      "epoch": 5.008974358974359,
+      "grad_norm": 37.676414489746094,
+      "learning_rate": 5.897435897435898e-06,
+      "loss": 1.9409,
+      "step": 1380
+    },
+    {
+      "epoch": 5.009401709401709,
+      "grad_norm": 0.4700198471546173,
+      "learning_rate": 5.940170940170941e-06,
+      "loss": 1.4757,
+      "step": 1390
+    },
+    {
+      "epoch": 5.00982905982906,
+      "grad_norm": 0.603115975856781,
+      "learning_rate": 5.982905982905983e-06,
+      "loss": 1.4646,
+      "step": 1400
+    },
+    {
+      "epoch": 5.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 3.5354690551757812,
+      "eval_runtime": 33.4888,
+      "eval_samples_per_second": 0.747,
+      "eval_steps_per_second": 0.747,
+      "step": 1404
+    },
+    {
+      "epoch": 6.00025641025641,
+      "grad_norm": 0.20566977560520172,
+      "learning_rate": 6.025641025641026e-06,
+      "loss": 2.5308,
+      "step": 1410
+    },
+    {
+      "epoch": 6.000683760683761,
+      "grad_norm": 1.3430235385894775,
+      "learning_rate": 6.0683760683760684e-06,
+      "loss": 1.7984,
+      "step": 1420
+    },
+    {
+      "epoch": 6.001111111111111,
+      "grad_norm": 34.01142120361328,
+      "learning_rate": 6.111111111111112e-06,
+      "loss": 0.5602,
+      "step": 1430
+    },
+    {
+      "epoch": 6.001538461538462,
+      "grad_norm": 0.08811581879854202,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.5698,
+      "step": 1440
+    },
+    {
+      "epoch": 6.001965811965812,
+      "grad_norm": 31.072364807128906,
+      "learning_rate": 6.196581196581198e-06,
+      "loss": 1.9642,
+      "step": 1450
+    },
+    {
+      "epoch": 6.002393162393163,
+      "grad_norm": 0.3049279749393463,
+      "learning_rate": 6.23931623931624e-06,
+      "loss": 1.0951,
+      "step": 1460
+    },
+    {
+      "epoch": 6.002820512820513,
+      "grad_norm": 1.261183500289917,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 1.5718,
+      "step": 1470
+    },
+    {
+      "epoch": 6.003247863247863,
+      "grad_norm": 29.76000213623047,
+      "learning_rate": 6.324786324786325e-06,
+      "loss": 2.3047,
+      "step": 1480
+    },
+    {
+      "epoch": 6.003675213675214,
+      "grad_norm": 0.3626205325126648,
+      "learning_rate": 6.367521367521368e-06,
+      "loss": 1.7411,
+      "step": 1490
+    },
+    {
+      "epoch": 6.004102564102564,
+      "grad_norm": 27.959877014160156,
+      "learning_rate": 6.410256410256412e-06,
+      "loss": 2.4126,
+      "step": 1500
+    },
+    {
+      "epoch": 6.004529914529915,
+      "grad_norm": 0.5188838243484497,
+      "learning_rate": 6.4529914529914535e-06,
+      "loss": 1.3862,
+      "step": 1510
+    },
+    {
+      "epoch": 6.004957264957265,
+      "grad_norm": 0.29587042331695557,
+      "learning_rate": 6.495726495726496e-06,
+      "loss": 1.414,
+      "step": 1520
+    },
+    {
+      "epoch": 6.0053846153846155,
+      "grad_norm": 0.25534921884536743,
+      "learning_rate": 6.538461538461539e-06,
+      "loss": 1.7212,
+      "step": 1530
+    },
+    {
+      "epoch": 6.005811965811966,
+      "grad_norm": 0.2307765781879425,
+      "learning_rate": 6.581196581196582e-06,
+      "loss": 2.294,
+      "step": 1540
+    },
+    {
+      "epoch": 6.006239316239316,
+      "grad_norm": 28.337125778198242,
+      "learning_rate": 6.623931623931624e-06,
+      "loss": 0.946,
+      "step": 1550
+    },
+    {
+      "epoch": 6.006666666666667,
+      "grad_norm": 0.16784150898456573,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.1255,
+      "step": 1560
+    },
+    {
+      "epoch": 6.007094017094017,
+      "grad_norm": 27.95771598815918,
+      "learning_rate": 6.7094017094017094e-06,
+      "loss": 1.6833,
+      "step": 1570
+    },
+    {
+      "epoch": 6.007521367521368,
+      "grad_norm": 0.4080459773540497,
+      "learning_rate": 6.752136752136753e-06,
+      "loss": 1.3406,
+      "step": 1580
+    },
+    {
+      "epoch": 6.007948717948718,
+      "grad_norm": 0.8113716244697571,
+      "learning_rate": 6.794871794871796e-06,
+      "loss": 2.6112,
+      "step": 1590
+    },
+    {
+      "epoch": 6.0083760683760685,
+      "grad_norm": 30.46502685546875,
+      "learning_rate": 6.837606837606839e-06,
+      "loss": 1.4458,
+      "step": 1600
+    },
+    {
+      "epoch": 6.008803418803419,
+      "grad_norm": 0.22283227741718292,
+      "learning_rate": 6.880341880341881e-06,
+      "loss": 0.5283,
+      "step": 1610
+    },
+    {
+      "epoch": 6.009230769230769,
+      "grad_norm": 28.292848587036133,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 1.5511,
+      "step": 1620
+    },
+    {
+      "epoch": 6.00965811965812,
+      "grad_norm": 28.0618839263916,
+      "learning_rate": 6.965811965811966e-06,
+      "loss": 2.1201,
+      "step": 1630
+    },
+    {
+      "epoch": 6.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 3.0478899478912354,
+      "eval_runtime": 33.4068,
+      "eval_samples_per_second": 0.748,
+      "eval_steps_per_second": 0.748,
+      "step": 1638
+    },
+    {
+      "epoch": 7.00008547008547,
+      "grad_norm": 33.30900573730469,
+      "learning_rate": 7.008547008547009e-06,
+      "loss": 1.429,
+      "step": 1640
+    },
+    {
+      "epoch": 7.000512820512821,
+      "grad_norm": 29.006980895996094,
+      "learning_rate": 7.051282051282053e-06,
+      "loss": 1.6062,
+      "step": 1650
+    },
+    {
+      "epoch": 7.000940170940171,
+      "grad_norm": 28.740236282348633,
+      "learning_rate": 7.0940170940170945e-06,
+      "loss": 2.1064,
+      "step": 1660
+    },
+    {
+      "epoch": 7.001367521367522,
+      "grad_norm": 0.5548082590103149,
+      "learning_rate": 7.136752136752137e-06,
+      "loss": 1.4227,
+      "step": 1670
+    },
+    {
+      "epoch": 7.001794871794872,
+      "grad_norm": 0.2548494040966034,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 0.4677,
+      "step": 1680
+    },
+    {
+      "epoch": 7.002222222222223,
+      "grad_norm": 0.08594454079866409,
+      "learning_rate": 7.222222222222223e-06,
+      "loss": 0.6487,
+      "step": 1690
+    },
+    {
+      "epoch": 7.002649572649573,
+      "grad_norm": 0.11181553453207016,
+      "learning_rate": 7.264957264957266e-06,
+      "loss": 0.6856,
+      "step": 1700
+    },
+    {
+      "epoch": 7.003076923076923,
+      "grad_norm": 0.09680427610874176,
+      "learning_rate": 7.307692307692308e-06,
+      "loss": 1.2589,
+      "step": 1710
+    },
+    {
+      "epoch": 7.003504273504274,
+      "grad_norm": 0.18477901816368103,
+      "learning_rate": 7.350427350427351e-06,
+      "loss": 2.0357,
+      "step": 1720
+    },
+    {
+      "epoch": 7.003931623931624,
+      "grad_norm": 28.30834197998047,
+      "learning_rate": 7.393162393162394e-06,
+      "loss": 1.7978,
+      "step": 1730
+    },
+    {
+      "epoch": 7.004358974358975,
+      "grad_norm": 0.4138849675655365,
+      "learning_rate": 7.435897435897437e-06,
+      "loss": 1.4938,
+      "step": 1740
+    },
+    {
+      "epoch": 7.004786324786325,
+      "grad_norm": 46.25049591064453,
+      "learning_rate": 7.47863247863248e-06,
+      "loss": 1.2354,
+      "step": 1750
+    },
+    {
+      "epoch": 7.0052136752136756,
+      "grad_norm": 0.2702569365501404,
+      "learning_rate": 7.521367521367522e-06,
+      "loss": 1.8753,
+      "step": 1760
+    },
+    {
+      "epoch": 7.005641025641026,
+      "grad_norm": 0.2681766450405121,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 1.2446,
+      "step": 1770
+    },
+    {
+      "epoch": 7.006068376068376,
+      "grad_norm": 47.34070587158203,
+      "learning_rate": 7.606837606837607e-06,
+      "loss": 2.2493,
+      "step": 1780
+    },
+    {
+      "epoch": 7.006495726495727,
+      "grad_norm": 31.515134811401367,
+      "learning_rate": 7.649572649572649e-06,
+      "loss": 2.1708,
+      "step": 1790
+    },
+    {
+      "epoch": 7.006923076923077,
+      "grad_norm": 0.44051307439804077,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 0.9425,
+      "step": 1800
+    },
+    {
+      "epoch": 7.007350427350428,
+      "grad_norm": 0.16471922397613525,
+      "learning_rate": 7.735042735042736e-06,
+      "loss": 1.6227,
+      "step": 1810
+    },
+    {
+      "epoch": 7.007777777777778,
+      "grad_norm": 36.21978759765625,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 3.136,
+      "step": 1820
+    },
+    {
+      "epoch": 7.0082051282051285,
+      "grad_norm": 0.6311964988708496,
+      "learning_rate": 7.820512820512822e-06,
+      "loss": 1.6591,
+      "step": 1830
+    },
+    {
+      "epoch": 7.008632478632479,
+      "grad_norm": 0.15079300105571747,
+      "learning_rate": 7.863247863247863e-06,
+      "loss": 0.4923,
+      "step": 1840
+    },
+    {
+      "epoch": 7.009059829059829,
+      "grad_norm": 0.0793166533112526,
+      "learning_rate": 7.905982905982906e-06,
+      "loss": 0.6274,
+      "step": 1850
+    },
+    {
+      "epoch": 7.00948717948718,
+      "grad_norm": 39.972232818603516,
+      "learning_rate": 7.948717948717949e-06,
+      "loss": 1.902,
+      "step": 1860
+    },
+    {
+      "epoch": 7.00991452991453,
+      "grad_norm": 2.84613299369812,
+      "learning_rate": 7.991452991452993e-06,
+      "loss": 2.9021,
+      "step": 1870
+    },
+    {
+      "epoch": 7.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 2.8181352615356445,
+      "eval_runtime": 33.4387,
+      "eval_samples_per_second": 0.748,
+      "eval_steps_per_second": 0.748,
+      "step": 1872
+    },
+    {
+      "epoch": 8.00034188034188,
+      "grad_norm": 32.629676818847656,
+      "learning_rate": 8.034188034188036e-06,
+      "loss": 0.6487,
+      "step": 1880
+    },
+    {
+      "epoch": 8.000769230769231,
+      "grad_norm": 0.23920656740665436,
+      "learning_rate": 8.076923076923077e-06,
+      "loss": 1.2006,
+      "step": 1890
+    },
+    {
+      "epoch": 8.00119658119658,
+      "grad_norm": 0.14234939217567444,
+      "learning_rate": 8.11965811965812e-06,
+      "loss": 1.2065,
+      "step": 1900
+    },
+    {
+      "epoch": 8.001623931623932,
+      "grad_norm": 0.12460368871688843,
+      "learning_rate": 8.162393162393163e-06,
+      "loss": 0.5436,
+      "step": 1910
+    },
+    {
+      "epoch": 8.002051282051282,
+      "grad_norm": 0.43331626057624817,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 1.5192,
+      "step": 1920
+    },
+    {
+      "epoch": 8.002478632478633,
+      "grad_norm": 0.12986142933368683,
+      "learning_rate": 8.247863247863248e-06,
+      "loss": 0.6905,
+      "step": 1930
+    },
+    {
+      "epoch": 8.002905982905983,
+      "grad_norm": 0.5832127928733826,
+      "learning_rate": 8.290598290598293e-06,
+      "loss": 3.0564,
+      "step": 1940
+    },
+    {
+      "epoch": 8.003333333333334,
+      "grad_norm": 28.78748321533203,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 2.3034,
+      "step": 1950
+    },
+    {
+      "epoch": 8.003760683760683,
+      "grad_norm": 31.847166061401367,
+      "learning_rate": 8.376068376068377e-06,
+      "loss": 1.9432,
+      "step": 1960
+    },
+    {
+      "epoch": 8.004188034188035,
+      "grad_norm": 0.23129618167877197,
+      "learning_rate": 8.41880341880342e-06,
+      "loss": 0.8151,
+      "step": 1970
+    },
+    {
+      "epoch": 8.004615384615384,
+      "grad_norm": 0.08719867467880249,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 0.4525,
+      "step": 1980
+    },
+    {
+      "epoch": 8.005042735042736,
+      "grad_norm": 0.17510618269443512,
+      "learning_rate": 8.504273504273505e-06,
+      "loss": 2.3329,
+      "step": 1990
+    },
+    {
+      "epoch": 8.005470085470085,
+      "grad_norm": 0.23968863487243652,
+      "learning_rate": 8.547008547008548e-06,
+      "loss": 1.4327,
+      "step": 2000
+    },
+    {
+      "epoch": 8.005897435897436,
+      "grad_norm": 0.1258096992969513,
+      "learning_rate": 8.58974358974359e-06,
+      "loss": 1.5496,
+      "step": 2010
+    },
+    {
+      "epoch": 8.006324786324786,
+      "grad_norm": 33.13344192504883,
+      "learning_rate": 8.632478632478633e-06,
+      "loss": 1.8633,
+      "step": 2020
+    },
+    {
+      "epoch": 8.006752136752137,
+      "grad_norm": 34.57917022705078,
+      "learning_rate": 8.675213675213676e-06,
+      "loss": 2.1156,
+      "step": 2030
+    },
+    {
+      "epoch": 8.007179487179487,
+      "grad_norm": 30.499267578125,
+      "learning_rate": 8.717948717948719e-06,
+      "loss": 1.4366,
+      "step": 2040
+    },
+    {
+      "epoch": 8.007606837606838,
+      "grad_norm": 0.2296626716852188,
+      "learning_rate": 8.760683760683762e-06,
+      "loss": 0.5388,
+      "step": 2050
+    },
+    {
+      "epoch": 8.008034188034188,
+      "grad_norm": 0.6007127165794373,
+      "learning_rate": 8.803418803418804e-06,
+      "loss": 2.0019,
+      "step": 2060
+    },
+    {
+      "epoch": 8.008461538461539,
+      "grad_norm": 29.225831985473633,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 1.6262,
+      "step": 2070
+    },
+    {
+      "epoch": 8.008888888888889,
+      "grad_norm": 0.9478847980499268,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 2.0313,
+      "step": 2080
+    },
+    {
+      "epoch": 8.00931623931624,
+      "grad_norm": 0.12494786083698273,
+      "learning_rate": 8.931623931623933e-06,
+      "loss": 0.9103,
+      "step": 2090
+    },
+    {
+      "epoch": 8.00974358974359,
+      "grad_norm": 23.248950958251953,
+      "learning_rate": 8.974358974358976e-06,
+      "loss": 2.1527,
+      "step": 2100
+    },
+    {
+      "epoch": 8.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 2.7605271339416504,
+      "eval_runtime": 32.6542,
+      "eval_samples_per_second": 0.766,
+      "eval_steps_per_second": 0.766,
+      "step": 2106
+    },
+    {
+      "epoch": 9.00017094017094,
+      "grad_norm": 25.770069122314453,
+      "learning_rate": 9.017094017094018e-06,
+      "loss": 1.5706,
+      "step": 2110
+    },
+    {
+      "epoch": 9.00059829059829,
+      "grad_norm": 0.7022117376327515,
+      "learning_rate": 9.059829059829061e-06,
+      "loss": 0.4205,
+      "step": 2120
+    },
+    {
+      "epoch": 9.001025641025642,
+      "grad_norm": 0.08867479860782623,
+      "learning_rate": 9.102564102564104e-06,
+      "loss": 1.2611,
+      "step": 2130
+    },
+    {
+      "epoch": 9.001452991452991,
+      "grad_norm": 0.09678523242473602,
+      "learning_rate": 9.145299145299145e-06,
+      "loss": 1.1773,
+      "step": 2140
+    },
+    {
+      "epoch": 9.001880341880343,
+      "grad_norm": 0.15343564748764038,
+      "learning_rate": 9.188034188034188e-06,
+      "loss": 1.2586,
+      "step": 2150
+    },
+    {
+      "epoch": 9.002307692307692,
+      "grad_norm": 0.35261672735214233,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.8426,
+      "step": 2160
+    },
+    {
+      "epoch": 9.002735042735043,
+      "grad_norm": 0.2280280739068985,
+      "learning_rate": 9.273504273504275e-06,
+      "loss": 1.1276,
+      "step": 2170
+    },
+    {
+      "epoch": 9.003162393162393,
+      "grad_norm": 0.11199548840522766,
+      "learning_rate": 9.316239316239318e-06,
+      "loss": 2.7832,
+      "step": 2180
+    },
+    {
+      "epoch": 9.003589743589744,
+      "grad_norm": 0.2621850073337555,
+      "learning_rate": 9.358974358974359e-06,
+      "loss": 0.9458,
+      "step": 2190
+    },
+    {
+      "epoch": 9.004017094017094,
+      "grad_norm": 0.36012938618659973,
+      "learning_rate": 9.401709401709402e-06,
+      "loss": 1.1242,
+      "step": 2200
+    },
+    {
+      "epoch": 9.004444444444445,
+      "grad_norm": 48.78242492675781,
+      "learning_rate": 9.444444444444445e-06,
+      "loss": 3.3251,
+      "step": 2210
+    },
+    {
+      "epoch": 9.004871794871795,
+      "grad_norm": 0.15066808462142944,
+      "learning_rate": 9.487179487179487e-06,
+      "loss": 1.708,
+      "step": 2220
+    },
+    {
+      "epoch": 9.005299145299146,
+      "grad_norm": 0.4691851735115051,
+      "learning_rate": 9.52991452991453e-06,
+      "loss": 1.553,
+      "step": 2230
+    },
+    {
+      "epoch": 9.005726495726496,
+      "grad_norm": 28.600677490234375,
+      "learning_rate": 9.572649572649575e-06,
+      "loss": 1.2926,
+      "step": 2240
+    },
+    {
+      "epoch": 9.006153846153847,
+      "grad_norm": 32.24127960205078,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 2.4733,
+      "step": 2250
+    },
+    {
+      "epoch": 9.006581196581196,
+      "grad_norm": 0.7631069421768188,
+      "learning_rate": 9.658119658119659e-06,
+      "loss": 1.7091,
+      "step": 2260
+    },
+    {
+      "epoch": 9.007008547008548,
+      "grad_norm": 0.2995574176311493,
+      "learning_rate": 9.700854700854701e-06,
+      "loss": 2.0053,
+      "step": 2270
+    },
+    {
+      "epoch": 9.007435897435897,
+      "grad_norm": 3.1298446655273438,
+      "learning_rate": 9.743589743589744e-06,
+      "loss": 1.1043,
+      "step": 2280
+    },
+    {
+      "epoch": 9.007863247863249,
+      "grad_norm": 0.1203409880399704,
+      "learning_rate": 9.786324786324787e-06,
+      "loss": 0.6468,
+      "step": 2290
+    },
+    {
+      "epoch": 9.008290598290598,
+      "grad_norm": 0.3753611445426941,
+      "learning_rate": 9.82905982905983e-06,
+      "loss": 1.8003,
+      "step": 2300
+    },
+    {
+      "epoch": 9.00871794871795,
+      "grad_norm": 0.24749663472175598,
+      "learning_rate": 9.871794871794872e-06,
+      "loss": 0.6302,
+      "step": 2310
+    },
+    {
+      "epoch": 9.009145299145299,
+      "grad_norm": 0.15980985760688782,
+      "learning_rate": 9.914529914529915e-06,
+      "loss": 1.7506,
+      "step": 2320
+    },
+    {
+      "epoch": 9.00957264957265,
+      "grad_norm": 30.294530868530273,
+      "learning_rate": 9.957264957264958e-06,
+      "loss": 2.5406,
+      "step": 2330
+    },
+    {
+      "epoch": 9.01,
+      "grad_norm": 1.435558557510376,
+      "learning_rate": 1e-05,
+      "loss": 1.9428,
+      "step": 2340
+    },
+    {
+      "epoch": 9.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 2.4512882232666016,
+      "eval_runtime": 32.6431,
+      "eval_samples_per_second": 0.766,
+      "eval_steps_per_second": 0.766,
+      "step": 2340
+    },
+    {
+      "epoch": 10.000427350427351,
+      "grad_norm": 26.23455810546875,
+      "learning_rate": 9.99525166191833e-06,
+      "loss": 0.9502,
+      "step": 2350
+    },
+    {
+      "epoch": 10.0008547008547,
+      "grad_norm": 28.52158546447754,
+      "learning_rate": 9.990503323836657e-06,
+      "loss": 1.7874,
+      "step": 2360
+    },
+    {
+      "epoch": 10.001282051282052,
+      "grad_norm": 0.10940665751695633,
+      "learning_rate": 9.985754985754987e-06,
+      "loss": 1.1145,
+      "step": 2370
+    },
+    {
+      "epoch": 10.001709401709402,
+      "grad_norm": 0.49054867029190063,
+      "learning_rate": 9.981006647673314e-06,
+      "loss": 2.1067,
+      "step": 2380
+    },
+    {
+      "epoch": 10.002136752136753,
+      "grad_norm": 0.4312061071395874,
+      "learning_rate": 9.976258309591643e-06,
+      "loss": 1.4459,
+      "step": 2390
+    },
+    {
+      "epoch": 10.002564102564103,
+      "grad_norm": 0.9604039788246155,
+      "learning_rate": 9.971509971509972e-06,
+      "loss": 1.5741,
+      "step": 2400
+    },
+    {
+      "epoch": 10.002991452991454,
+      "grad_norm": 28.289756774902344,
+      "learning_rate": 9.966761633428301e-06,
+      "loss": 3.0157,
+      "step": 2410
+    },
+    {
+      "epoch": 10.003418803418803,
+      "grad_norm": 1.3765814304351807,
+      "learning_rate": 9.96201329534663e-06,
+      "loss": 1.7118,
+      "step": 2420
+    },
+    {
+      "epoch": 10.003846153846155,
+      "grad_norm": 0.4429996609687805,
+      "learning_rate": 9.957264957264958e-06,
+      "loss": 1.8346,
+      "step": 2430
+    },
+    {
+      "epoch": 10.004273504273504,
+      "grad_norm": 0.1344616711139679,
+      "learning_rate": 9.952516619183287e-06,
+      "loss": 1.1072,
+      "step": 2440
+    },
+    {
+      "epoch": 10.004700854700856,
+      "grad_norm": 0.2778800129890442,
+      "learning_rate": 9.947768281101615e-06,
+      "loss": 1.0562,
+      "step": 2450
+    },
+    {
+      "epoch": 10.005128205128205,
+      "grad_norm": 41.122074127197266,
+      "learning_rate": 9.943019943019944e-06,
+      "loss": 2.096,
+      "step": 2460
+    },
+    {
+      "epoch": 10.005555555555556,
+      "grad_norm": 0.6196286082267761,
+      "learning_rate": 9.938271604938273e-06,
+      "loss": 1.9945,
+      "step": 2470
+    },
+    {
+      "epoch": 10.005982905982906,
+      "grad_norm": 38.4301872253418,
+      "learning_rate": 9.9335232668566e-06,
+      "loss": 1.5745,
+      "step": 2480
+    },
+    {
+      "epoch": 10.006410256410257,
+      "grad_norm": 0.5608100295066833,
+      "learning_rate": 9.92877492877493e-06,
+      "loss": 1.7862,
+      "step": 2490
+    },
+    {
+      "epoch": 10.006837606837607,
+      "grad_norm": 0.08879340440034866,
+      "learning_rate": 9.924026590693259e-06,
+      "loss": 1.2266,
+      "step": 2500
+    },
+    {
+      "epoch": 10.007264957264958,
+      "grad_norm": 0.14399921894073486,
+      "learning_rate": 9.919278252611588e-06,
+      "loss": 1.5128,
+      "step": 2510
+    },
+    {
+      "epoch": 10.007692307692308,
+      "grad_norm": 0.567003071308136,
+      "learning_rate": 9.914529914529915e-06,
+      "loss": 0.5539,
+      "step": 2520
+    },
+    {
+      "epoch": 10.008119658119659,
+      "grad_norm": 0.1469346433877945,
+      "learning_rate": 9.909781576448244e-06,
+      "loss": 1.3608,
+      "step": 2530
+    },
+    {
+      "epoch": 10.008547008547009,
+      "grad_norm": 0.6318036913871765,
+      "learning_rate": 9.905033238366572e-06,
+      "loss": 0.6639,
+      "step": 2540
+    },
+    {
+      "epoch": 10.00897435897436,
+      "grad_norm": 0.13676932454109192,
+      "learning_rate": 9.900284900284901e-06,
+      "loss": 2.7342,
+      "step": 2550
+    },
+    {
+      "epoch": 10.00940170940171,
+      "grad_norm": 28.927698135375977,
+      "learning_rate": 9.89553656220323e-06,
+      "loss": 2.5179,
+      "step": 2560
+    },
+    {
+      "epoch": 10.00982905982906,
+      "grad_norm": 26.96585464477539,
+      "learning_rate": 9.890788224121558e-06,
+      "loss": 1.6949,
+      "step": 2570
+    },
+    {
+      "epoch": 10.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 3.2310287952423096,
+      "eval_runtime": 32.7403,
+      "eval_samples_per_second": 0.764,
+      "eval_steps_per_second": 0.764,
+      "step": 2574
+    },
+    {
+      "epoch": 11.00025641025641,
+      "grad_norm": 29.56972312927246,
+      "learning_rate": 9.886039886039887e-06,
+      "loss": 2.0711,
+      "step": 2580
+    },
+    {
+      "epoch": 11.00068376068376,
+      "grad_norm": 1.1823328733444214,
+      "learning_rate": 9.881291547958214e-06,
+      "loss": 1.3947,
+      "step": 2590
+    },
+    {
+      "epoch": 11.001111111111111,
+      "grad_norm": 34.95638656616211,
+      "learning_rate": 9.876543209876543e-06,
+      "loss": 1.3976,
+      "step": 2600
+    },
+    {
+      "epoch": 11.00153846153846,
+      "grad_norm": 0.5684574842453003,
+      "learning_rate": 9.871794871794872e-06,
+      "loss": 0.5828,
+      "step": 2610
+    },
+    {
+      "epoch": 11.001965811965812,
+      "grad_norm": 0.13000904023647308,
+      "learning_rate": 9.867046533713202e-06,
+      "loss": 0.547,
+      "step": 2620
+    },
+    {
+      "epoch": 11.002393162393162,
+      "grad_norm": 0.08917149901390076,
+      "learning_rate": 9.86229819563153e-06,
+      "loss": 2.5733,
+      "step": 2630
+    },
+    {
+      "epoch": 11.002820512820513,
+      "grad_norm": 27.73358154296875,
+      "learning_rate": 9.857549857549858e-06,
+      "loss": 3.1395,
+      "step": 2640
+    },
+    {
+      "epoch": 11.003247863247863,
+      "grad_norm": 24.41839027404785,
+      "learning_rate": 9.852801519468187e-06,
+      "loss": 1.2438,
+      "step": 2650
+    },
+    {
+      "epoch": 11.003675213675214,
+      "grad_norm": 3.812605142593384,
+      "learning_rate": 9.848053181386515e-06,
+      "loss": 1.9698,
+      "step": 2660
+    },
+    {
+      "epoch": 11.004102564102563,
+      "grad_norm": 1.3905110359191895,
+      "learning_rate": 9.843304843304844e-06,
+      "loss": 1.0108,
+      "step": 2670
+    },
+    {
+      "epoch": 11.004529914529915,
+      "grad_norm": 0.7647677063941956,
+      "learning_rate": 9.838556505223173e-06,
+      "loss": 2.3101,
+      "step": 2680
+    },
+    {
+      "epoch": 11.004957264957264,
+      "grad_norm": 31.24732780456543,
+      "learning_rate": 9.8338081671415e-06,
+      "loss": 2.3872,
+      "step": 2690
+    },
+    {
+      "epoch": 11.005384615384616,
+      "grad_norm": 29.57086944580078,
+      "learning_rate": 9.82905982905983e-06,
+      "loss": 1.7839,
+      "step": 2700
+    },
+    {
+      "epoch": 11.005811965811965,
+      "grad_norm": 0.18307341635227203,
+      "learning_rate": 9.824311490978159e-06,
+      "loss": 1.4724,
+      "step": 2710
+    },
+    {
+      "epoch": 11.006239316239316,
+      "grad_norm": 0.0853017047047615,
+      "learning_rate": 9.819563152896488e-06,
+      "loss": 0.2625,
+      "step": 2720
+    },
+    {
+      "epoch": 11.006666666666666,
+      "grad_norm": 0.10152238607406616,
+      "learning_rate": 9.814814814814815e-06,
+      "loss": 1.0233,
+      "step": 2730
+    },
+    {
+      "epoch": 11.007094017094017,
+      "grad_norm": 4.792705535888672,
+      "learning_rate": 9.810066476733145e-06,
+      "loss": 2.28,
+      "step": 2740
+    },
+    {
+      "epoch": 11.007521367521367,
+      "grad_norm": 0.5689181685447693,
+      "learning_rate": 9.805318138651474e-06,
+      "loss": 1.3607,
+      "step": 2750
+    },
+    {
+      "epoch": 11.007948717948718,
+      "grad_norm": 0.07235058397054672,
+      "learning_rate": 9.800569800569801e-06,
+      "loss": 0.5546,
+      "step": 2760
+    },
+    {
+      "epoch": 11.008376068376068,
+      "grad_norm": 0.18848180770874023,
+      "learning_rate": 9.79582146248813e-06,
+      "loss": 0.0045,
+      "step": 2770
+    },
+    {
+      "epoch": 11.008803418803419,
+      "grad_norm": 0.14690843224525452,
+      "learning_rate": 9.791073124406458e-06,
+      "loss": 0.6331,
+      "step": 2780
+    },
+    {
+      "epoch": 11.009230769230768,
+      "grad_norm": 35.11496353149414,
+      "learning_rate": 9.786324786324787e-06,
+      "loss": 1.3858,
+      "step": 2790
+    },
+    {
+      "epoch": 11.00965811965812,
+      "grad_norm": 0.35278671979904175,
+      "learning_rate": 9.781576448243116e-06,
+      "loss": 0.7839,
+      "step": 2800
+    },
+    {
+      "epoch": 11.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 3.237220525741577,
+      "eval_runtime": 31.3652,
+      "eval_samples_per_second": 0.797,
+      "eval_steps_per_second": 0.797,
+      "step": 2808
+    },
+    {
+      "epoch": 12.00008547008547,
+      "grad_norm": 0.31852149963378906,
+      "learning_rate": 9.776828110161445e-06,
+      "loss": 1.8794,
+      "step": 2810
+    },
+    {
+      "epoch": 12.00051282051282,
+      "grad_norm": 1.0761709213256836,
+      "learning_rate": 9.772079772079773e-06,
+      "loss": 1.9101,
+      "step": 2820
+    },
+    {
+      "epoch": 12.00094017094017,
+      "grad_norm": 0.06357964873313904,
+      "learning_rate": 9.767331433998102e-06,
+      "loss": 1.1383,
+      "step": 2830
+    },
+    {
+      "epoch": 12.001367521367522,
+      "grad_norm": 54.093082427978516,
+      "learning_rate": 9.762583095916431e-06,
+      "loss": 2.112,
+      "step": 2840
+    },
+    {
+      "epoch": 12.001794871794871,
+      "grad_norm": 0.2823314070701599,
+      "learning_rate": 9.757834757834758e-06,
+      "loss": 0.5963,
+      "step": 2850
+    },
+    {
+      "epoch": 12.002222222222223,
+      "grad_norm": 0.3854888677597046,
+      "learning_rate": 9.753086419753087e-06,
+      "loss": 0.5196,
+      "step": 2860
+    },
+    {
+      "epoch": 12.002649572649572,
+      "grad_norm": 32.18788146972656,
+      "learning_rate": 9.748338081671415e-06,
+      "loss": 3.0031,
+      "step": 2870
+    },
+    {
+      "epoch": 12.003076923076923,
+      "grad_norm": 1.2660186290740967,
+      "learning_rate": 9.743589743589744e-06,
+      "loss": 1.9453,
+      "step": 2880
+    },
+    {
+      "epoch": 12.003504273504273,
+      "grad_norm": 49.34856033325195,
+      "learning_rate": 9.738841405508073e-06,
+      "loss": 2.1836,
+      "step": 2890
+    },
+    {
+      "epoch": 12.003931623931624,
+      "grad_norm": 3.923487663269043,
+      "learning_rate": 9.7340930674264e-06,
+      "loss": 0.6615,
+      "step": 2900
+    },
+    {
+      "epoch": 12.004358974358974,
+      "grad_norm": 0.09915890544652939,
+      "learning_rate": 9.72934472934473e-06,
+      "loss": 1.8817,
+      "step": 2910
+    },
+    {
+      "epoch": 12.004786324786325,
+      "grad_norm": 0.21953581273555756,
+      "learning_rate": 9.724596391263059e-06,
+      "loss": 2.0783,
+      "step": 2920
+    },
+    {
+      "epoch": 12.005213675213675,
+      "grad_norm": 0.0524006113409996,
+      "learning_rate": 9.719848053181388e-06,
+      "loss": 1.1279,
+      "step": 2930
+    },
+    {
+      "epoch": 12.005641025641026,
+      "grad_norm": 56.099334716796875,
+      "learning_rate": 9.715099715099716e-06,
+      "loss": 1.4809,
+      "step": 2940
+    },
+    {
+      "epoch": 12.006068376068376,
+      "grad_norm": 0.5895985960960388,
+      "learning_rate": 9.710351377018045e-06,
+      "loss": 0.5329,
+      "step": 2950
+    },
+    {
+      "epoch": 12.006495726495727,
+      "grad_norm": 42.8546257019043,
+      "learning_rate": 9.705603038936374e-06,
+      "loss": 2.1514,
+      "step": 2960
+    },
+    {
+      "epoch": 12.006923076923076,
+      "grad_norm": 0.03076835907995701,
+      "learning_rate": 9.700854700854701e-06,
+      "loss": 0.4221,
+      "step": 2970
+    },
+    {
+      "epoch": 12.007350427350428,
+      "grad_norm": 0.02532465010881424,
+      "learning_rate": 9.69610636277303e-06,
+      "loss": 2.3099,
+      "step": 2980
+    },
+    {
+      "epoch": 12.007777777777777,
+      "grad_norm": 62.39934158325195,
+      "learning_rate": 9.691358024691358e-06,
+      "loss": 0.6646,
+      "step": 2990
+    },
+    {
+      "epoch": 12.008205128205129,
+      "grad_norm": 54.54960632324219,
+      "learning_rate": 9.686609686609687e-06,
+      "loss": 0.4507,
+      "step": 3000
+    },
+    {
+      "epoch": 12.008632478632478,
+      "grad_norm": 1.0829317569732666,
+      "learning_rate": 9.681861348528016e-06,
+      "loss": 1.9857,
+      "step": 3010
+    },
+    {
+      "epoch": 12.00905982905983,
+      "grad_norm": 11.404279708862305,
+      "learning_rate": 9.677113010446345e-06,
+      "loss": 1.1354,
+      "step": 3020
+    },
+    {
+      "epoch": 12.009487179487179,
+      "grad_norm": 0.041183874011039734,
+      "learning_rate": 9.672364672364673e-06,
+      "loss": 0.8741,
+      "step": 3030
+    },
+    {
+      "epoch": 12.00991452991453,
+      "grad_norm": 19.812788009643555,
+      "learning_rate": 9.667616334283002e-06,
+      "loss": 0.3228,
+      "step": 3040
+    },
+    {
+      "epoch": 12.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 4.458773136138916,
+      "eval_runtime": 31.367,
+      "eval_samples_per_second": 0.797,
+      "eval_steps_per_second": 0.797,
+      "step": 3042
+    },
+    {
+      "epoch": 13.00034188034188,
+      "grad_norm": 49.4742317199707,
+      "learning_rate": 9.662867996201331e-06,
+      "loss": 2.7291,
+      "step": 3050
+    },
+    {
+      "epoch": 13.000769230769231,
+      "grad_norm": 33.5203857421875,
+      "learning_rate": 9.658119658119659e-06,
+      "loss": 1.3205,
+      "step": 3060
+    },
+    {
+      "epoch": 13.00119658119658,
+      "grad_norm": 32.33980941772461,
+      "learning_rate": 9.653371320037988e-06,
+      "loss": 1.9122,
+      "step": 3070
+    },
+    {
+      "epoch": 13.001623931623932,
+      "grad_norm": 0.2694286108016968,
+      "learning_rate": 9.648622981956317e-06,
+      "loss": 1.0847,
+      "step": 3080
+    },
+    {
+      "epoch": 13.002051282051282,
+      "grad_norm": 0.17534177005290985,
+      "learning_rate": 9.643874643874644e-06,
+      "loss": 1.2066,
+      "step": 3090
+    },
+    {
+      "epoch": 13.002478632478633,
+      "grad_norm": 0.06266115605831146,
+      "learning_rate": 9.639126305792973e-06,
+      "loss": 1.0707,
+      "step": 3100
+    },
+    {
+      "epoch": 13.002905982905983,
+      "grad_norm": 28.773160934448242,
+      "learning_rate": 9.6343779677113e-06,
+      "loss": 3.3066,
+      "step": 3110
+    },
+    {
+      "epoch": 13.003333333333334,
+      "grad_norm": 36.008384704589844,
+      "learning_rate": 9.62962962962963e-06,
+      "loss": 1.0375,
+      "step": 3120
+    },
+    {
+      "epoch": 13.003760683760683,
+      "grad_norm": 0.03540613874793053,
+      "learning_rate": 9.624881291547959e-06,
+      "loss": 0.5259,
+      "step": 3130
+    },
+    {
+      "epoch": 13.004188034188035,
+      "grad_norm": 0.03709403797984123,
+      "learning_rate": 9.620132953466288e-06,
+      "loss": 0.0029,
+      "step": 3140
+    },
+    {
+      "epoch": 13.004615384615384,
+      "grad_norm": 0.5573925375938416,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 1.7656,
+      "step": 3150
+    },
+    {
+      "epoch": 13.005042735042736,
+      "grad_norm": 0.03247790411114693,
+      "learning_rate": 9.610636277302945e-06,
+      "loss": 1.6948,
+      "step": 3160
+    },
+    {
+      "epoch": 13.005470085470085,
+      "grad_norm": 32.801788330078125,
+      "learning_rate": 9.605887939221274e-06,
+      "loss": 2.2156,
+      "step": 3170
+    },
+    {
+      "epoch": 13.005897435897436,
+      "grad_norm": 3.8869693279266357,
+      "learning_rate": 9.601139601139601e-06,
+      "loss": 1.0666,
+      "step": 3180
+    },
+    {
+      "epoch": 13.006324786324786,
+      "grad_norm": 0.8103746175765991,
+      "learning_rate": 9.59639126305793e-06,
+      "loss": 1.5307,
+      "step": 3190
+    },
+    {
+      "epoch": 13.006752136752137,
+      "grad_norm": 10.717965126037598,
+      "learning_rate": 9.591642924976258e-06,
+      "loss": 1.6417,
+      "step": 3200
+    },
+    {
+      "epoch": 13.007179487179487,
+      "grad_norm": 41.972896575927734,
+      "learning_rate": 9.586894586894587e-06,
+      "loss": 0.5114,
+      "step": 3210
+    },
+    {
+      "epoch": 13.007606837606838,
+      "grad_norm": 79.40288543701172,
+      "learning_rate": 9.582146248812916e-06,
+      "loss": 1.7693,
+      "step": 3220
+    },
+    {
+      "epoch": 13.008034188034188,
+      "grad_norm": 0.05152087286114693,
+      "learning_rate": 9.577397910731245e-06,
+      "loss": 1.1784,
+      "step": 3230
+    },
+    {
+      "epoch": 13.008461538461539,
+      "grad_norm": 0.10308440029621124,
+      "learning_rate": 9.572649572649575e-06,
+      "loss": 1.9728,
+      "step": 3240
+    },
+    {
+      "epoch": 13.008888888888889,
+      "grad_norm": 0.10380467772483826,
+      "learning_rate": 9.567901234567902e-06,
+      "loss": 1.1541,
+      "step": 3250
+    },
+    {
+      "epoch": 13.00931623931624,
+      "grad_norm": 0.8398118019104004,
+      "learning_rate": 9.563152896486231e-06,
+      "loss": 0.5554,
+      "step": 3260
+    },
+    {
+      "epoch": 13.00974358974359,
+      "grad_norm": 41.57830810546875,
+      "learning_rate": 9.558404558404559e-06,
+      "loss": 3.5377,
+      "step": 3270
+    },
+    {
+      "epoch": 13.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 2.862149238586426,
+      "eval_runtime": 31.9079,
+      "eval_samples_per_second": 0.784,
+      "eval_steps_per_second": 0.784,
+      "step": 3276
+    },
+    {
+      "epoch": 14.00017094017094,
+      "grad_norm": 37.766448974609375,
+      "learning_rate": 9.553656220322888e-06,
+      "loss": 1.0101,
+      "step": 3280
+    },
+    {
+      "epoch": 14.00059829059829,
+      "grad_norm": 0.0569000206887722,
+      "learning_rate": 9.548907882241217e-06,
+      "loss": 0.5743,
+      "step": 3290
+    },
+    {
+      "epoch": 14.001025641025642,
+      "grad_norm": 0.07739593833684921,
+      "learning_rate": 9.544159544159544e-06,
+      "loss": 0.6754,
+      "step": 3300
+    },
+    {
+      "epoch": 14.001452991452991,
+      "grad_norm": 0.08230680227279663,
+      "learning_rate": 9.539411206077874e-06,
+      "loss": 2.4253,
+      "step": 3310
+    },
+    {
+      "epoch": 14.001880341880343,
+      "grad_norm": 64.62518310546875,
+      "learning_rate": 9.534662867996201e-06,
+      "loss": 2.6924,
+      "step": 3320
+    },
+    {
+      "epoch": 14.002307692307692,
+      "grad_norm": 6.906940460205078,
+      "learning_rate": 9.52991452991453e-06,
+      "loss": 1.5079,
+      "step": 3330
+    },
+    {
+      "epoch": 14.002735042735043,
+      "grad_norm": 42.747108459472656,
+      "learning_rate": 9.52516619183286e-06,
+      "loss": 0.4776,
+      "step": 3340
+    },
+    {
+      "epoch": 14.003162393162393,
+      "grad_norm": 0.0985599160194397,
+      "learning_rate": 9.520417853751188e-06,
+      "loss": 1.2448,
+      "step": 3350
+    },
+    {
+      "epoch": 14.003589743589744,
+      "grad_norm": 43.92565155029297,
+      "learning_rate": 9.515669515669516e-06,
+      "loss": 1.1799,
+      "step": 3360
+    },
+    {
+      "epoch": 14.004017094017094,
+      "grad_norm": 0.21025124192237854,
+      "learning_rate": 9.510921177587845e-06,
+      "loss": 1.5539,
+      "step": 3370
+    },
+    {
+      "epoch": 14.004444444444445,
+      "grad_norm": 0.03208040073513985,
+      "learning_rate": 9.506172839506174e-06,
+      "loss": 0.0046,
+      "step": 3380
+    },
+    {
+      "epoch": 14.004871794871795,
+      "grad_norm": 0.054694049060344696,
+      "learning_rate": 9.501424501424502e-06,
+      "loss": 1.6939,
+      "step": 3390
+    },
+    {
+      "epoch": 14.005299145299146,
+      "grad_norm": 70.33654022216797,
+      "learning_rate": 9.49667616334283e-06,
+      "loss": 1.1934,
+      "step": 3400
+    },
+    {
+      "epoch": 14.005726495726496,
+      "grad_norm": 22.107633590698242,
+      "learning_rate": 9.49192782526116e-06,
+      "loss": 1.3934,
+      "step": 3410
+    },
+    {
+      "epoch": 14.006153846153847,
+      "grad_norm": 24.67310905456543,
+      "learning_rate": 9.487179487179487e-06,
+      "loss": 1.545,
+      "step": 3420
+    },
+    {
+      "epoch": 14.006581196581196,
+      "grad_norm": 15.487519264221191,
+      "learning_rate": 9.482431149097816e-06,
+      "loss": 1.0724,
+      "step": 3430
+    },
+    {
+      "epoch": 14.007008547008548,
+      "grad_norm": 8.748615264892578,
+      "learning_rate": 9.477682811016146e-06,
+      "loss": 1.1913,
+      "step": 3440
+    },
+    {
+      "epoch": 14.007435897435897,
+      "grad_norm": 0.04509786143898964,
+      "learning_rate": 9.472934472934475e-06,
+      "loss": 0.8465,
+      "step": 3450
+    },
+    {
+      "epoch": 14.007863247863249,
+      "grad_norm": 0.052408602088689804,
+      "learning_rate": 9.468186134852802e-06,
+      "loss": 0.9335,
+      "step": 3460
+    },
+    {
+      "epoch": 14.008290598290598,
+      "grad_norm": 4.355685234069824,
+      "learning_rate": 9.463437796771131e-06,
+      "loss": 1.4667,
+      "step": 3470
+    },
+    {
+      "epoch": 14.00871794871795,
+      "grad_norm": 31.14105796813965,
+      "learning_rate": 9.458689458689459e-06,
+      "loss": 1.9456,
+      "step": 3480
+    },
+    {
+      "epoch": 14.009145299145299,
+      "grad_norm": 5.5146284103393555,
+      "learning_rate": 9.453941120607788e-06,
+      "loss": 2.4975,
+      "step": 3490
+    },
+    {
+      "epoch": 14.00957264957265,
+      "grad_norm": 1.885633945465088,
+      "learning_rate": 9.449192782526117e-06,
+      "loss": 1.6884,
+      "step": 3500
+    },
+    {
+      "epoch": 14.01,
+      "grad_norm": 0.19068482518196106,
+      "learning_rate": 9.444444444444445e-06,
+      "loss": 0.509,
+      "step": 3510
+    },
+    {
+      "epoch": 14.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 2.746034860610962,
+      "eval_runtime": 31.4372,
+      "eval_samples_per_second": 0.795,
+      "eval_steps_per_second": 0.795,
+      "step": 3510
+    },
+    {
+      "epoch": 15.000427350427351,
+      "grad_norm": 0.03531758114695549,
+      "learning_rate": 9.439696106362774e-06,
+      "loss": 1.7623,
+      "step": 3520
+    },
+    {
+      "epoch": 15.0008547008547,
+      "grad_norm": 0.2500402331352234,
+      "learning_rate": 9.434947768281101e-06,
+      "loss": 0.9918,
+      "step": 3530
+    },
+    {
+      "epoch": 15.001282051282052,
+      "grad_norm": 6.20103120803833,
+      "learning_rate": 9.430199430199432e-06,
+      "loss": 1.3489,
+      "step": 3540
+    },
+    {
+      "epoch": 15.001709401709402,
+      "grad_norm": 0.275524377822876,
+      "learning_rate": 9.42545109211776e-06,
+      "loss": 0.6611,
+      "step": 3550
+    },
+    {
+      "epoch": 15.002136752136753,
+      "grad_norm": 0.08374079316854477,
+      "learning_rate": 9.420702754036089e-06,
+      "loss": 1.2316,
+      "step": 3560
+    },
+    {
+      "epoch": 15.002564102564103,
+      "grad_norm": 0.5503836274147034,
+      "learning_rate": 9.415954415954418e-06,
+      "loss": 0.8269,
+      "step": 3570
+    },
+    {
+      "epoch": 15.002991452991454,
+      "grad_norm": 46.557186126708984,
+      "learning_rate": 9.411206077872745e-06,
+      "loss": 1.2204,
+      "step": 3580
+    },
+    {
+      "epoch": 15.003418803418803,
+      "grad_norm": 63.70065689086914,
+      "learning_rate": 9.406457739791074e-06,
+      "loss": 1.834,
+      "step": 3590
+    },
+    {
+      "epoch": 15.003846153846155,
+      "grad_norm": 35.3852424621582,
+      "learning_rate": 9.401709401709402e-06,
+      "loss": 0.6711,
+      "step": 3600
+    },
+    {
+      "epoch": 15.004273504273504,
+      "grad_norm": 16.211149215698242,
+      "learning_rate": 9.396961063627731e-06,
+      "loss": 0.6182,
+      "step": 3610
+    },
+    {
+      "epoch": 15.004700854700856,
+      "grad_norm": 76.63227081298828,
+      "learning_rate": 9.39221272554606e-06,
+      "loss": 0.831,
+      "step": 3620
+    },
+    {
+      "epoch": 15.005128205128205,
+      "grad_norm": 0.5662261843681335,
+      "learning_rate": 9.387464387464388e-06,
+      "loss": 0.6825,
+      "step": 3630
+    },
+    {
+      "epoch": 15.005555555555556,
+      "grad_norm": 0.5306532382965088,
+      "learning_rate": 9.382716049382717e-06,
+      "loss": 1.213,
+      "step": 3640
+    },
+    {
+      "epoch": 15.005982905982906,
+      "grad_norm": 74.40409088134766,
+      "learning_rate": 9.377967711301046e-06,
+      "loss": 0.5874,
+      "step": 3650
+    },
+    {
+      "epoch": 15.006410256410257,
+      "grad_norm": 0.02636171504855156,
+      "learning_rate": 9.373219373219375e-06,
+      "loss": 0.6467,
+      "step": 3660
+    },
+    {
+      "epoch": 15.006837606837607,
+      "grad_norm": 0.043768566101789474,
+      "learning_rate": 9.368471035137702e-06,
+      "loss": 1.444,
+      "step": 3670
+    },
+    {
+      "epoch": 15.007264957264958,
+      "grad_norm": 1.168853998184204,
+      "learning_rate": 9.363722697056032e-06,
+      "loss": 2.6958,
+      "step": 3680
+    },
+    {
+      "epoch": 15.007692307692308,
+      "grad_norm": 41.314117431640625,
+      "learning_rate": 9.358974358974359e-06,
+      "loss": 1.917,
+      "step": 3690
+    },
+    {
+      "epoch": 15.008119658119659,
+      "grad_norm": 28.175498962402344,
+      "learning_rate": 9.354226020892688e-06,
+      "loss": 1.4498,
+      "step": 3700
+    },
+    {
+      "epoch": 15.008547008547009,
+      "grad_norm": 0.07334843277931213,
+      "learning_rate": 9.349477682811017e-06,
+      "loss": 0.3526,
+      "step": 3710
+    },
+    {
+      "epoch": 15.00897435897436,
+      "grad_norm": 0.06012887507677078,
+      "learning_rate": 9.344729344729345e-06,
+      "loss": 1.0099,
+      "step": 3720
+    },
+    {
+      "epoch": 15.00940170940171,
+      "grad_norm": 0.030554110184311867,
+      "learning_rate": 9.339981006647674e-06,
+      "loss": 0.5927,
+      "step": 3730
+    },
+    {
+      "epoch": 15.00982905982906,
+      "grad_norm": 0.12558279931545258,
+      "learning_rate": 9.335232668566003e-06,
+      "loss": 0.1437,
+      "step": 3740
+    },
+    {
+      "epoch": 15.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 2.969764471054077,
+      "eval_runtime": 31.3741,
+      "eval_samples_per_second": 0.797,
+      "eval_steps_per_second": 0.797,
+      "step": 3744
+    },
+    {
+      "epoch": 16.00025641025641,
+      "grad_norm": 0.020290929824113846,
+      "learning_rate": 9.330484330484332e-06,
+      "loss": 1.4002,
+      "step": 3750
+    },
+    {
+      "epoch": 16.00068376068376,
+      "grad_norm": 0.02340000309050083,
+      "learning_rate": 9.32573599240266e-06,
+      "loss": 1.3384,
+      "step": 3760
+    },
+    {
+      "epoch": 16.00111111111111,
+      "grad_norm": 0.027016054838895798,
+      "learning_rate": 9.320987654320989e-06,
+      "loss": 2.2665,
+      "step": 3770
+    },
+    {
+      "epoch": 16.001538461538463,
+      "grad_norm": 0.11080523580312729,
+      "learning_rate": 9.316239316239318e-06,
+      "loss": 0.7879,
+      "step": 3780
+    },
+    {
+      "epoch": 16.00196581196581,
+      "grad_norm": 123.3178939819336,
+      "learning_rate": 9.311490978157645e-06,
+      "loss": 1.1529,
+      "step": 3790
+    },
+    {
+      "epoch": 16.00239316239316,
+      "grad_norm": 0.09909241646528244,
+      "learning_rate": 9.306742640075974e-06,
+      "loss": 2.4706,
+      "step": 3800
+    },
+    {
+      "epoch": 16.002820512820513,
+      "grad_norm": 0.025457249954342842,
+      "learning_rate": 9.301994301994302e-06,
+      "loss": 0.1416,
+      "step": 3810
+    },
+    {
+      "epoch": 16.003247863247864,
+      "grad_norm": 0.054812464863061905,
+      "learning_rate": 9.297245963912631e-06,
+      "loss": 0.1454,
+      "step": 3820
+    },
+    {
+      "epoch": 16.003675213675212,
+      "grad_norm": 42.573997497558594,
+      "learning_rate": 9.29249762583096e-06,
+      "loss": 2.6915,
+      "step": 3830
+    },
+    {
+      "epoch": 16.004102564102563,
+      "grad_norm": 0.10259363800287247,
+      "learning_rate": 9.287749287749288e-06,
+      "loss": 1.2003,
+      "step": 3840
+    },
+    {
+      "epoch": 16.004529914529915,
+      "grad_norm": 0.0471661239862442,
+      "learning_rate": 9.283000949667617e-06,
+      "loss": 1.2392,
+      "step": 3850
+    },
+    {
+      "epoch": 16.004957264957266,
+      "grad_norm": 0.0715036392211914,
+      "learning_rate": 9.278252611585946e-06,
+      "loss": 0.4382,
+      "step": 3860
+    },
+    {
+      "epoch": 16.005384615384614,
+      "grad_norm": 0.045701488852500916,
+      "learning_rate": 9.273504273504275e-06,
+      "loss": 0.3563,
+      "step": 3870
+    },
+    {
+      "epoch": 16.005811965811965,
+      "grad_norm": 91.8859634399414,
+      "learning_rate": 9.268755935422603e-06,
+      "loss": 1.0391,
+      "step": 3880
+    },
+    {
+      "epoch": 16.006239316239316,
+      "grad_norm": 2.5818240642547607,
+      "learning_rate": 9.264007597340932e-06,
+      "loss": 1.1685,
+      "step": 3890
+    },
+    {
+      "epoch": 16.006666666666668,
+      "grad_norm": 0.031820375472307205,
+      "learning_rate": 9.25925925925926e-06,
+      "loss": 1.1833,
+      "step": 3900
+    },
+    {
+      "epoch": 16.007094017094015,
+      "grad_norm": 0.07577929645776749,
+      "learning_rate": 9.254510921177588e-06,
+      "loss": 2.5019,
+      "step": 3910
+    },
+    {
+      "epoch": 16.007521367521367,
+      "grad_norm": 0.03320784494280815,
+      "learning_rate": 9.249762583095917e-06,
+      "loss": 0.9758,
+      "step": 3920
+    },
+    {
+      "epoch": 16.007948717948718,
+      "grad_norm": 117.20964813232422,
+      "learning_rate": 9.245014245014245e-06,
+      "loss": 0.7272,
+      "step": 3930
+    },
+    {
+      "epoch": 16.00837606837607,
+      "grad_norm": 12.37725830078125,
+      "learning_rate": 9.240265906932574e-06,
+      "loss": 0.4982,
+      "step": 3940
+    },
+    {
+      "epoch": 16.008803418803417,
+      "grad_norm": 0.03360183909535408,
+      "learning_rate": 9.235517568850903e-06,
+      "loss": 1.0855,
+      "step": 3950
+    },
+    {
+      "epoch": 16.00923076923077,
+      "grad_norm": 209.72621154785156,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 1.1257,
+      "step": 3960
+    },
+    {
+      "epoch": 16.00965811965812,
+      "grad_norm": 45.13322830200195,
+      "learning_rate": 9.22602089268756e-06,
+      "loss": 1.0039,
+      "step": 3970
+    },
+    {
+      "epoch": 16.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 1.9414650201797485,
+      "eval_runtime": 31.717,
+      "eval_samples_per_second": 0.788,
+      "eval_steps_per_second": 0.788,
+      "step": 3978
+    },
+    {
+      "epoch": 17.00008547008547,
+      "grad_norm": 0.20372648537158966,
+      "learning_rate": 9.221272554605889e-06,
+      "loss": 0.989,
+      "step": 3980
+    },
+    {
+      "epoch": 17.00051282051282,
+      "grad_norm": 143.5639190673828,
+      "learning_rate": 9.216524216524218e-06,
+      "loss": 0.5736,
+      "step": 3990
+    },
+    {
+      "epoch": 17.000940170940172,
+      "grad_norm": 0.0925712063908577,
+      "learning_rate": 9.211775878442545e-06,
+      "loss": 0.7445,
+      "step": 4000
+    },
+    {
+      "epoch": 17.00136752136752,
+      "grad_norm": 0.03829040750861168,
+      "learning_rate": 9.207027540360875e-06,
+      "loss": 0.5319,
+      "step": 4010
+    },
+    {
+      "epoch": 17.00179487179487,
+      "grad_norm": 66.21209716796875,
+      "learning_rate": 9.202279202279202e-06,
+      "loss": 1.2316,
+      "step": 4020
+    },
+    {
+      "epoch": 17.002222222222223,
+      "grad_norm": 0.07042970508337021,
+      "learning_rate": 9.197530864197531e-06,
+      "loss": 1.7251,
+      "step": 4030
+    },
+    {
+      "epoch": 17.002649572649574,
+      "grad_norm": 1.4991275072097778,
+      "learning_rate": 9.19278252611586e-06,
+      "loss": 0.9513,
+      "step": 4040
+    },
+    {
+      "epoch": 17.00307692307692,
+      "grad_norm": 0.5810855627059937,
+      "learning_rate": 9.188034188034188e-06,
+      "loss": 2.5822,
+      "step": 4050
+    },
+    {
+      "epoch": 17.003504273504273,
+      "grad_norm": 0.08620700240135193,
+      "learning_rate": 9.183285849952517e-06,
+      "loss": 1.3987,
+      "step": 4060
+    },
+    {
+      "epoch": 17.003931623931624,
+      "grad_norm": 35.556392669677734,
+      "learning_rate": 9.178537511870846e-06,
+      "loss": 1.2302,
+      "step": 4070
+    },
+    {
+      "epoch": 17.004358974358976,
+      "grad_norm": 33.437049865722656,
+      "learning_rate": 9.173789173789175e-06,
+      "loss": 1.1155,
+      "step": 4080
+    },
+    {
+      "epoch": 17.004786324786323,
+      "grad_norm": 0.6656836867332458,
+      "learning_rate": 9.169040835707503e-06,
+      "loss": 1.1672,
+      "step": 4090
+    },
+    {
+      "epoch": 17.005213675213675,
+      "grad_norm": 0.0775466188788414,
+      "learning_rate": 9.164292497625832e-06,
+      "loss": 0.9781,
+      "step": 4100
+    },
+    {
+      "epoch": 17.005641025641026,
+      "grad_norm": 177.47276306152344,
+      "learning_rate": 9.159544159544161e-06,
+      "loss": 1.6938,
+      "step": 4110
+    },
+    {
+      "epoch": 17.006068376068377,
+      "grad_norm": 0.02730586938560009,
+      "learning_rate": 9.154795821462488e-06,
+      "loss": 1.2777,
+      "step": 4120
+    },
+    {
+      "epoch": 17.006495726495725,
+      "grad_norm": 76.6297607421875,
+      "learning_rate": 9.150047483380818e-06,
+      "loss": 0.6365,
+      "step": 4130
+    },
+    {
+      "epoch": 17.006923076923076,
+      "grad_norm": 1.5181001424789429,
+      "learning_rate": 9.145299145299145e-06,
+      "loss": 0.5115,
+      "step": 4140
+    },
+    {
+      "epoch": 17.007350427350428,
+      "grad_norm": 7.319108486175537,
+      "learning_rate": 9.140550807217474e-06,
+      "loss": 1.4807,
+      "step": 4150
+    },
+    {
+      "epoch": 17.00777777777778,
+      "grad_norm": 139.0987548828125,
+      "learning_rate": 9.135802469135803e-06,
+      "loss": 1.5217,
+      "step": 4160
+    },
+    {
+      "epoch": 17.008205128205127,
+      "grad_norm": 16.966917037963867,
+      "learning_rate": 9.131054131054132e-06,
+      "loss": 2.0712,
+      "step": 4170
+    },
+    {
+      "epoch": 17.008632478632478,
+      "grad_norm": 73.7410659790039,
+      "learning_rate": 9.126305792972462e-06,
+      "loss": 1.0309,
+      "step": 4180
+    },
+    {
+      "epoch": 17.00905982905983,
+      "grad_norm": 0.03598838672041893,
+      "learning_rate": 9.121557454890789e-06,
+      "loss": 0.3113,
+      "step": 4190
+    },
+    {
+      "epoch": 17.00948717948718,
+      "grad_norm": 0.01311536505818367,
+      "learning_rate": 9.116809116809118e-06,
+      "loss": 1.8983,
+      "step": 4200
+    },
+    {
+      "epoch": 17.00991452991453,
+      "grad_norm": 0.07452392578125,
+      "learning_rate": 9.112060778727446e-06,
+      "loss": 0.0062,
+      "step": 4210
+    },
+    {
+      "epoch": 17.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 3.7040703296661377,
+      "eval_runtime": 31.9027,
+      "eval_samples_per_second": 0.784,
+      "eval_steps_per_second": 0.784,
+      "step": 4212
+    },
+    {
+      "epoch": 18.00034188034188,
+      "grad_norm": 107.44599151611328,
+      "learning_rate": 9.107312440645775e-06,
+      "loss": 1.8731,
+      "step": 4220
+    },
+    {
+      "epoch": 18.00076923076923,
+      "grad_norm": 0.23524224758148193,
+      "learning_rate": 9.102564102564104e-06,
+      "loss": 1.0308,
+      "step": 4230
+    },
+    {
+      "epoch": 18.00119658119658,
+      "grad_norm": 0.07270976901054382,
+      "learning_rate": 9.097815764482431e-06,
+      "loss": 1.2283,
+      "step": 4240
+    },
+    {
+      "epoch": 18.001623931623932,
+      "grad_norm": 6.627790451049805,
+      "learning_rate": 9.09306742640076e-06,
+      "loss": 1.0094,
+      "step": 4250
+    },
+    {
+      "epoch": 18.002051282051283,
+      "grad_norm": 0.02103256992995739,
+      "learning_rate": 9.088319088319088e-06,
+      "loss": 0.5875,
+      "step": 4260
+    },
+    {
+      "epoch": 18.00247863247863,
+      "grad_norm": 0.6851010322570801,
+      "learning_rate": 9.083570750237419e-06,
+      "loss": 1.2334,
+      "step": 4270
+    },
+    {
+      "epoch": 18.002905982905983,
+      "grad_norm": 82.27792358398438,
+      "learning_rate": 9.078822412155746e-06,
+      "loss": 1.9545,
+      "step": 4280
+    },
+    {
+      "epoch": 18.003333333333334,
+      "grad_norm": 139.23216247558594,
+      "learning_rate": 9.074074074074075e-06,
+      "loss": 0.8405,
+      "step": 4290
+    },
+    {
+      "epoch": 18.003760683760685,
+      "grad_norm": 0.8755741715431213,
+      "learning_rate": 9.069325735992403e-06,
+      "loss": 1.1443,
+      "step": 4300
+    },
+    {
+      "epoch": 18.004188034188033,
+      "grad_norm": 105.50877380371094,
+      "learning_rate": 9.064577397910732e-06,
+      "loss": 0.821,
+      "step": 4310
+    },
+    {
+      "epoch": 18.004615384615384,
+      "grad_norm": 114.59051513671875,
+      "learning_rate": 9.059829059829061e-06,
+      "loss": 1.6665,
+      "step": 4320
+    },
+    {
+      "epoch": 18.005042735042736,
+      "grad_norm": 0.04322560876607895,
+      "learning_rate": 9.055080721747389e-06,
+      "loss": 0.9944,
+      "step": 4330
+    },
+    {
+      "epoch": 18.005470085470087,
+      "grad_norm": 16.47235870361328,
+      "learning_rate": 9.050332383665718e-06,
+      "loss": 0.6319,
+      "step": 4340
+    },
+    {
+      "epoch": 18.005897435897435,
+      "grad_norm": 5.790500640869141,
+      "learning_rate": 9.045584045584045e-06,
+      "loss": 1.4753,
+      "step": 4350
+    },
+    {
+      "epoch": 18.006324786324786,
+      "grad_norm": 0.045558616518974304,
+      "learning_rate": 9.040835707502374e-06,
+      "loss": 1.4482,
+      "step": 4360
+    },
+    {
+      "epoch": 18.006752136752137,
+      "grad_norm": 0.014822369441390038,
+      "learning_rate": 9.036087369420703e-06,
+      "loss": 1.1214,
+      "step": 4370
+    },
+    {
+      "epoch": 18.00717948717949,
+      "grad_norm": 90.13900756835938,
+      "learning_rate": 9.031339031339033e-06,
+      "loss": 2.0725,
+      "step": 4380
+    },
+    {
+      "epoch": 18.007606837606836,
+      "grad_norm": 80.13975524902344,
+      "learning_rate": 9.026590693257362e-06,
+      "loss": 1.1801,
+      "step": 4390
+    },
+    {
+      "epoch": 18.008034188034188,
+      "grad_norm": 0.009048054926097393,
+      "learning_rate": 9.02184235517569e-06,
+      "loss": 0.6618,
+      "step": 4400
+    },
+    {
+      "epoch": 18.00846153846154,
+      "grad_norm": 59.612308502197266,
+      "learning_rate": 9.017094017094018e-06,
+      "loss": 1.7319,
+      "step": 4410
+    },
+    {
+      "epoch": 18.00888888888889,
+      "grad_norm": 0.1411258578300476,
+      "learning_rate": 9.012345679012346e-06,
+      "loss": 0.9388,
+      "step": 4420
+    },
+    {
+      "epoch": 18.009316239316238,
+      "grad_norm": 12.921591758728027,
+      "learning_rate": 9.007597340930675e-06,
+      "loss": 0.6901,
+      "step": 4430
+    },
+    {
+      "epoch": 18.00974358974359,
+      "grad_norm": 25.335466384887695,
+      "learning_rate": 9.002849002849004e-06,
+      "loss": 0.6038,
+      "step": 4440
+    },
+    {
+      "epoch": 18.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 3.2141363620758057,
+      "eval_runtime": 31.3241,
+      "eval_samples_per_second": 0.798,
+      "eval_steps_per_second": 0.798,
+      "step": 4446
+    },
+    {
+      "epoch": 19.00017094017094,
+      "grad_norm": 119.39568328857422,
+      "learning_rate": 8.998100664767332e-06,
+      "loss": 1.4807,
+      "step": 4450
+    },
+    {
+      "epoch": 19.00059829059829,
+      "grad_norm": 86.97232055664062,
+      "learning_rate": 8.99335232668566e-06,
+      "loss": 1.6098,
+      "step": 4460
+    },
+    {
+      "epoch": 19.00102564102564,
+      "grad_norm": 69.82275390625,
+      "learning_rate": 8.988603988603988e-06,
+      "loss": 0.7945,
+      "step": 4470
+    },
+    {
+      "epoch": 19.001452991452993,
+      "grad_norm": 0.024586718529462814,
+      "learning_rate": 8.983855650522319e-06,
+      "loss": 0.289,
+      "step": 4480
+    },
+    {
+      "epoch": 19.00188034188034,
+      "grad_norm": 0.08284483104944229,
+      "learning_rate": 8.979107312440646e-06,
+      "loss": 0.7986,
+      "step": 4490
+    },
+    {
+      "epoch": 19.002307692307692,
+      "grad_norm": 31.52510643005371,
+      "learning_rate": 8.974358974358976e-06,
+      "loss": 0.337,
+      "step": 4500
+    },
+    {
+      "epoch": 19.002735042735043,
+      "grad_norm": 0.04926073178648949,
+      "learning_rate": 8.969610636277305e-06,
+      "loss": 1.7346,
+      "step": 4510
+    },
+    {
+      "epoch": 19.003162393162395,
+      "grad_norm": 0.06321626901626587,
+      "learning_rate": 8.964862298195632e-06,
+      "loss": 0.9723,
+      "step": 4520
+    },
+    {
+      "epoch": 19.003589743589743,
+      "grad_norm": 0.09472198784351349,
+      "learning_rate": 8.960113960113961e-06,
+      "loss": 0.0057,
+      "step": 4530
+    },
+    {
+      "epoch": 19.004017094017094,
+      "grad_norm": 145.72950744628906,
+      "learning_rate": 8.955365622032289e-06,
+      "loss": 0.7365,
+      "step": 4540
+    },
+    {
+      "epoch": 19.004444444444445,
+      "grad_norm": 0.012398917227983475,
+      "learning_rate": 8.950617283950618e-06,
+      "loss": 2.1059,
+      "step": 4550
+    },
+    {
+      "epoch": 19.004871794871796,
+      "grad_norm": 73.18333435058594,
+      "learning_rate": 8.945868945868947e-06,
+      "loss": 1.8259,
+      "step": 4560
+    },
+    {
+      "epoch": 19.005299145299144,
+      "grad_norm": 84.90556335449219,
+      "learning_rate": 8.941120607787274e-06,
+      "loss": 0.696,
+      "step": 4570
+    },
+    {
+      "epoch": 19.005726495726496,
+      "grad_norm": 0.024428799748420715,
+      "learning_rate": 8.936372269705604e-06,
+      "loss": 1.0575,
+      "step": 4580
+    },
+    {
+      "epoch": 19.006153846153847,
+      "grad_norm": 44.86989974975586,
+      "learning_rate": 8.931623931623933e-06,
+      "loss": 1.0172,
+      "step": 4590
+    },
+    {
+      "epoch": 19.006581196581198,
+      "grad_norm": 0.16147296130657196,
+      "learning_rate": 8.926875593542262e-06,
+      "loss": 0.4637,
+      "step": 4600
+    },
+    {
+      "epoch": 19.007008547008546,
+      "grad_norm": 0.016476508229970932,
+      "learning_rate": 8.92212725546059e-06,
+      "loss": 0.0585,
+      "step": 4610
+    },
+    {
+      "epoch": 19.007435897435897,
+      "grad_norm": 69.78681945800781,
+      "learning_rate": 8.917378917378919e-06,
+      "loss": 2.8439,
+      "step": 4620
+    },
+    {
+      "epoch": 19.00786324786325,
+      "grad_norm": 143.7225799560547,
+      "learning_rate": 8.912630579297246e-06,
+      "loss": 0.3701,
+      "step": 4630
+    },
+    {
+      "epoch": 19.0082905982906,
+      "grad_norm": 3.918398141860962,
+      "learning_rate": 8.907882241215575e-06,
+      "loss": 0.6645,
+      "step": 4640
+    },
+    {
+      "epoch": 19.008717948717948,
+      "grad_norm": 0.4194408357143402,
+      "learning_rate": 8.903133903133904e-06,
+      "loss": 1.5778,
+      "step": 4650
+    },
+    {
+      "epoch": 19.0091452991453,
+      "grad_norm": 0.7064778804779053,
+      "learning_rate": 8.898385565052232e-06,
+      "loss": 0.9134,
+      "step": 4660
+    },
+    {
+      "epoch": 19.00957264957265,
+      "grad_norm": 97.42158508300781,
+      "learning_rate": 8.893637226970561e-06,
+      "loss": 0.6072,
+      "step": 4670
+    },
+    {
+      "epoch": 19.01,
+      "grad_norm": 0.028133299201726913,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 1.1687,
+      "step": 4680
+    },
+    {
+      "epoch": 19.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 2.407238721847534,
+      "eval_runtime": 31.5073,
+      "eval_samples_per_second": 0.793,
+      "eval_steps_per_second": 0.793,
+      "step": 4680
+    },
+    {
+      "epoch": 20.00042735042735,
+      "grad_norm": 0.0467204824090004,
+      "learning_rate": 8.884140550807219e-06,
+      "loss": 1.0038,
+      "step": 4690
+    },
+    {
+      "epoch": 20.000854700854703,
+      "grad_norm": 0.044175442308187485,
+      "learning_rate": 8.879392212725547e-06,
+      "loss": 0.2621,
+      "step": 4700
+    },
+    {
+      "epoch": 20.00128205128205,
+      "grad_norm": 0.12440741062164307,
+      "learning_rate": 8.874643874643876e-06,
+      "loss": 1.2026,
+      "step": 4710
+    },
+    {
+      "epoch": 20.0017094017094,
+      "grad_norm": 0.011806718073785305,
+      "learning_rate": 8.869895536562205e-06,
+      "loss": 0.7563,
+      "step": 4720
+    },
+    {
+      "epoch": 20.002136752136753,
+      "grad_norm": 1.0385252237319946,
+      "learning_rate": 8.865147198480532e-06,
+      "loss": 0.5996,
+      "step": 4730
+    },
+    {
+      "epoch": 20.002564102564104,
+      "grad_norm": 175.9308319091797,
+      "learning_rate": 8.860398860398861e-06,
+      "loss": 0.8629,
+      "step": 4740
+    },
+    {
+      "epoch": 20.002991452991452,
+      "grad_norm": 0.028463028371334076,
+      "learning_rate": 8.855650522317189e-06,
+      "loss": 2.8217,
+      "step": 4750
+    },
+    {
+      "epoch": 20.003418803418803,
+      "grad_norm": 0.030408434569835663,
+      "learning_rate": 8.850902184235518e-06,
+      "loss": 0.6615,
+      "step": 4760
+    },
+    {
+      "epoch": 20.003846153846155,
+      "grad_norm": 0.057413261383771896,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 0.5805,
+      "step": 4770
+    },
+    {
+      "epoch": 20.004273504273506,
+      "grad_norm": 73.27574920654297,
+      "learning_rate": 8.841405508072175e-06,
+      "loss": 1.1753,
+      "step": 4780
+    },
+    {
+      "epoch": 20.004700854700854,
+      "grad_norm": 0.06163284182548523,
+      "learning_rate": 8.836657169990504e-06,
+      "loss": 0.4274,
+      "step": 4790
+    },
+    {
+      "epoch": 20.005128205128205,
+      "grad_norm": 23.880149841308594,
+      "learning_rate": 8.831908831908833e-06,
+      "loss": 0.6001,
+      "step": 4800
+    },
+    {
+      "epoch": 20.005555555555556,
+      "grad_norm": 81.80984497070312,
+      "learning_rate": 8.827160493827162e-06,
+      "loss": 0.7434,
+      "step": 4810
+    },
+    {
+      "epoch": 20.005982905982908,
+      "grad_norm": 0.018306629732251167,
+      "learning_rate": 8.82241215574549e-06,
+      "loss": 0.7265,
+      "step": 4820
+    },
+    {
+      "epoch": 20.006410256410255,
+      "grad_norm": 2.1246113777160645,
+      "learning_rate": 8.817663817663819e-06,
+      "loss": 1.1732,
+      "step": 4830
+    },
+    {
+      "epoch": 20.006837606837607,
+      "grad_norm": 0.1039893627166748,
+      "learning_rate": 8.812915479582148e-06,
+      "loss": 1.2742,
+      "step": 4840
+    },
+    {
+      "epoch": 20.007264957264958,
+      "grad_norm": 0.3535962402820587,
+      "learning_rate": 8.808167141500475e-06,
+      "loss": 0.4528,
+      "step": 4850
+    },
+    {
+      "epoch": 20.00769230769231,
+      "grad_norm": 177.94131469726562,
+      "learning_rate": 8.803418803418804e-06,
+      "loss": 0.4491,
+      "step": 4860
+    },
+    {
+      "epoch": 20.008119658119657,
+      "grad_norm": 0.01753688044846058,
+      "learning_rate": 8.798670465337132e-06,
+      "loss": 1.9356,
+      "step": 4870
+    },
+    {
+      "epoch": 20.00854700854701,
+      "grad_norm": 5.777415752410889,
+      "learning_rate": 8.793922127255461e-06,
+      "loss": 0.7365,
+      "step": 4880
+    },
+    {
+      "epoch": 20.00897435897436,
+      "grad_norm": 0.021493086591362953,
+      "learning_rate": 8.78917378917379e-06,
+      "loss": 1.5335,
+      "step": 4890
+    },
+    {
+      "epoch": 20.00940170940171,
+      "grad_norm": 56.53275680541992,
+      "learning_rate": 8.78442545109212e-06,
+      "loss": 2.5917,
+      "step": 4900
+    },
+    {
+      "epoch": 20.00982905982906,
+      "grad_norm": 0.8556663393974304,
+      "learning_rate": 8.779677113010447e-06,
+      "loss": 0.8397,
+      "step": 4910
+    },
+    {
+      "epoch": 20.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 3.42124080657959,
+      "eval_runtime": 32.6346,
+      "eval_samples_per_second": 0.766,
+      "eval_steps_per_second": 0.766,
+      "step": 4914
+    },
+    {
+      "epoch": 21.00025641025641,
+      "grad_norm": 0.02766931988298893,
+      "learning_rate": 8.774928774928776e-06,
+      "loss": 1.0177,
+      "step": 4920
+    },
+    {
+      "epoch": 21.00068376068376,
+      "grad_norm": 1.8544671535491943,
+      "learning_rate": 8.770180436847105e-06,
+      "loss": 0.6464,
+      "step": 4930
+    },
+    {
+      "epoch": 21.00111111111111,
+      "grad_norm": 64.42668914794922,
+      "learning_rate": 8.765432098765432e-06,
+      "loss": 2.2753,
+      "step": 4940
+    },
+    {
+      "epoch": 21.001538461538463,
+      "grad_norm": 5.441823959350586,
+      "learning_rate": 8.760683760683762e-06,
+      "loss": 0.3229,
+      "step": 4950
+    },
+    {
+      "epoch": 21.00196581196581,
+      "grad_norm": 0.053660500794649124,
+      "learning_rate": 8.755935422602089e-06,
+      "loss": 0.4219,
+      "step": 4960
+    },
+    {
+      "epoch": 21.00239316239316,
+      "grad_norm": 0.015398544259369373,
+      "learning_rate": 8.751187084520418e-06,
+      "loss": 0.0334,
+      "step": 4970
+    },
+    {
+      "epoch": 21.002820512820513,
+      "grad_norm": 27.52494239807129,
+      "learning_rate": 8.746438746438747e-06,
+      "loss": 1.3975,
+      "step": 4980
+    },
+    {
+      "epoch": 21.003247863247864,
+      "grad_norm": 0.019361304119229317,
+      "learning_rate": 8.741690408357075e-06,
+      "loss": 1.1835,
+      "step": 4990
+    },
+    {
+      "epoch": 21.003675213675212,
+      "grad_norm": 0.15508921444416046,
+      "learning_rate": 8.736942070275406e-06,
+      "loss": 0.9132,
+      "step": 5000
+    },
+    {
+      "epoch": 21.004102564102563,
+      "grad_norm": 394.57867431640625,
+      "learning_rate": 8.732193732193733e-06,
+      "loss": 0.8003,
+      "step": 5010
+    },
+    {
+      "epoch": 21.004529914529915,
+      "grad_norm": 0.3072645962238312,
+      "learning_rate": 8.727445394112062e-06,
+      "loss": 1.085,
+      "step": 5020
+    },
+    {
+      "epoch": 21.004957264957266,
+      "grad_norm": 128.3138427734375,
+      "learning_rate": 8.72269705603039e-06,
+      "loss": 1.6623,
+      "step": 5030
+    },
+    {
+      "epoch": 21.005384615384614,
+      "grad_norm": 0.40651294589042664,
+      "learning_rate": 8.717948717948719e-06,
+      "loss": 1.3648,
+      "step": 5040
+    },
+    {
+      "epoch": 21.005811965811965,
+      "grad_norm": 51.56714630126953,
+      "learning_rate": 8.713200379867048e-06,
+      "loss": 1.1869,
+      "step": 5050
+    },
+    {
+      "epoch": 21.006239316239316,
+      "grad_norm": 167.30929565429688,
+      "learning_rate": 8.708452041785375e-06,
+      "loss": 1.2651,
+      "step": 5060
+    },
+    {
+      "epoch": 21.006666666666668,
+      "grad_norm": 26.427907943725586,
+      "learning_rate": 8.703703703703705e-06,
+      "loss": 1.136,
+      "step": 5070
+    },
+    {
+      "epoch": 21.007094017094015,
+      "grad_norm": 0.03196699172258377,
+      "learning_rate": 8.698955365622032e-06,
+      "loss": 0.5952,
+      "step": 5080
+    },
+    {
+      "epoch": 21.007521367521367,
+      "grad_norm": 131.19776916503906,
+      "learning_rate": 8.694207027540361e-06,
+      "loss": 1.0937,
+      "step": 5090
+    },
+    {
+      "epoch": 21.007948717948718,
+      "grad_norm": 0.009517078287899494,
+      "learning_rate": 8.68945868945869e-06,
+      "loss": 0.6601,
+      "step": 5100
+    },
+    {
+      "epoch": 21.00837606837607,
+      "grad_norm": 106.69688415527344,
+      "learning_rate": 8.68471035137702e-06,
+      "loss": 1.0542,
+      "step": 5110
+    },
+    {
+      "epoch": 21.008803418803417,
+      "grad_norm": 0.006986229680478573,
+      "learning_rate": 8.679962013295347e-06,
+      "loss": 1.1394,
+      "step": 5120
+    },
+    {
+      "epoch": 21.00923076923077,
+      "grad_norm": 127.98587799072266,
+      "learning_rate": 8.675213675213676e-06,
+      "loss": 2.0223,
+      "step": 5130
+    },
+    {
+      "epoch": 21.00965811965812,
+      "grad_norm": 113.3585205078125,
+      "learning_rate": 8.670465337132005e-06,
+      "loss": 1.1147,
+      "step": 5140
+    },
+    {
+      "epoch": 21.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 2.511547565460205,
+      "eval_runtime": 33.474,
+      "eval_samples_per_second": 0.747,
+      "eval_steps_per_second": 0.747,
+      "step": 5148
+    },
+    {
+      "epoch": 22.00008547008547,
+      "grad_norm": 0.017451312392950058,
+      "learning_rate": 8.665716999050333e-06,
+      "loss": 0.4376,
+      "step": 5150
+    },
+    {
+      "epoch": 22.00051282051282,
+      "grad_norm": 0.21703580021858215,
+      "learning_rate": 8.660968660968662e-06,
+      "loss": 0.8807,
+      "step": 5160
+    },
+    {
+      "epoch": 22.000940170940172,
+      "grad_norm": 0.01205768994987011,
+      "learning_rate": 8.656220322886991e-06,
+      "loss": 0.4462,
+      "step": 5170
+    },
+    {
+      "epoch": 22.00136752136752,
+      "grad_norm": 62.46580123901367,
+      "learning_rate": 8.651471984805318e-06,
+      "loss": 0.6433,
+      "step": 5180
+    },
+    {
+      "epoch": 22.00179487179487,
+      "grad_norm": 195.9849853515625,
+      "learning_rate": 8.646723646723648e-06,
+      "loss": 0.6842,
+      "step": 5190
+    },
+    {
+      "epoch": 22.002222222222223,
+      "grad_norm": 23.76949691772461,
+      "learning_rate": 8.641975308641975e-06,
+      "loss": 1.8101,
+      "step": 5200
+    },
+    {
+      "epoch": 22.002649572649574,
+      "grad_norm": 1.1113344430923462,
+      "learning_rate": 8.637226970560306e-06,
+      "loss": 1.3208,
+      "step": 5210
+    },
+    {
+      "epoch": 22.00307692307692,
+      "grad_norm": 50.05499267578125,
+      "learning_rate": 8.632478632478633e-06,
+      "loss": 1.7615,
+      "step": 5220
+    },
+    {
+      "epoch": 22.003504273504273,
+      "grad_norm": 148.08843994140625,
+      "learning_rate": 8.627730294396962e-06,
+      "loss": 0.5733,
+      "step": 5230
+    },
+    {
+      "epoch": 22.003931623931624,
+      "grad_norm": 22.12079429626465,
+      "learning_rate": 8.62298195631529e-06,
+      "loss": 0.399,
+      "step": 5240
+    },
+    {
+      "epoch": 22.004358974358976,
+      "grad_norm": 61.95510482788086,
+      "learning_rate": 8.618233618233619e-06,
+      "loss": 0.8791,
+      "step": 5250
+    },
+    {
+      "epoch": 22.004786324786323,
+      "grad_norm": 0.0063286456279456615,
+      "learning_rate": 8.613485280151948e-06,
+      "loss": 0.9967,
+      "step": 5260
+    },
+    {
+      "epoch": 22.005213675213675,
+      "grad_norm": 0.005704652518033981,
+      "learning_rate": 8.608736942070276e-06,
+      "loss": 0.0331,
+      "step": 5270
+    },
+    {
+      "epoch": 22.005641025641026,
+      "grad_norm": 0.019357306882739067,
+      "learning_rate": 8.603988603988605e-06,
+      "loss": 1.0326,
+      "step": 5280
+    },
+    {
+      "epoch": 22.006068376068377,
+      "grad_norm": 38.796791076660156,
+      "learning_rate": 8.599240265906932e-06,
+      "loss": 0.3044,
+      "step": 5290
+    },
+    {
+      "epoch": 22.006495726495725,
+      "grad_norm": 0.023437276482582092,
+      "learning_rate": 8.594491927825261e-06,
+      "loss": 1.0089,
+      "step": 5300
+    },
+    {
+      "epoch": 22.006923076923076,
+      "grad_norm": 91.38776397705078,
+      "learning_rate": 8.58974358974359e-06,
+      "loss": 1.7264,
+      "step": 5310
+    },
+    {
+      "epoch": 22.007350427350428,
+      "grad_norm": 0.006801048293709755,
+      "learning_rate": 8.58499525166192e-06,
+      "loss": 0.2886,
+      "step": 5320
+    },
+    {
+      "epoch": 22.00777777777778,
+      "grad_norm": 0.005525406915694475,
+      "learning_rate": 8.580246913580249e-06,
+      "loss": 0.6458,
+      "step": 5330
+    },
+    {
+      "epoch": 22.008205128205127,
+      "grad_norm": 396.8755187988281,
+      "learning_rate": 8.575498575498576e-06,
+      "loss": 0.6596,
+      "step": 5340
+    },
+    {
+      "epoch": 22.008632478632478,
+      "grad_norm": 0.45548421144485474,
+      "learning_rate": 8.570750237416905e-06,
+      "loss": 0.8506,
+      "step": 5350
+    },
+    {
+      "epoch": 22.00905982905983,
+      "grad_norm": 0.06096582114696503,
+      "learning_rate": 8.566001899335233e-06,
+      "loss": 0.9077,
+      "step": 5360
+    },
+    {
+      "epoch": 22.00948717948718,
+      "grad_norm": 52.80357360839844,
+      "learning_rate": 8.561253561253562e-06,
+      "loss": 1.4106,
+      "step": 5370
+    },
+    {
+      "epoch": 22.00991452991453,
+      "grad_norm": 0.40176790952682495,
+      "learning_rate": 8.556505223171891e-06,
+      "loss": 0.2286,
+      "step": 5380
+    },
+    {
+      "epoch": 22.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 2.434278964996338,
+      "eval_runtime": 33.4186,
+      "eval_samples_per_second": 0.748,
+      "eval_steps_per_second": 0.748,
+      "step": 5382
+    },
+    {
+      "epoch": 23.00034188034188,
+      "grad_norm": 0.008083014748990536,
+      "learning_rate": 8.551756885090219e-06,
+      "loss": 0.2302,
+      "step": 5390
+    },
+    {
+      "epoch": 23.00076923076923,
+      "grad_norm": 81.26300048828125,
+      "learning_rate": 8.547008547008548e-06,
+      "loss": 0.586,
+      "step": 5400
+    },
+    {
+      "epoch": 23.00119658119658,
+      "grad_norm": 0.00498996814712882,
+      "learning_rate": 8.542260208926875e-06,
+      "loss": 0.3741,
+      "step": 5410
+    },
+    {
+      "epoch": 23.001623931623932,
+      "grad_norm": 0.028612615540623665,
+      "learning_rate": 8.537511870845206e-06,
+      "loss": 0.4084,
+      "step": 5420
+    },
+    {
+      "epoch": 23.002051282051283,
+      "grad_norm": 0.005741656292229891,
+      "learning_rate": 8.532763532763533e-06,
+      "loss": 1.1549,
+      "step": 5430
+    },
+    {
+      "epoch": 23.00247863247863,
+      "grad_norm": 0.0038078909274190664,
+      "learning_rate": 8.528015194681863e-06,
+      "loss": 0.5803,
+      "step": 5440
+    },
+    {
+      "epoch": 23.002905982905983,
+      "grad_norm": 0.07777408510446548,
+      "learning_rate": 8.52326685660019e-06,
+      "loss": 1.8444,
+      "step": 5450
+    },
+    {
+      "epoch": 23.003333333333334,
+      "grad_norm": 0.23302805423736572,
+      "learning_rate": 8.518518518518519e-06,
+      "loss": 0.3736,
+      "step": 5460
+    },
+    {
+      "epoch": 23.003760683760685,
+      "grad_norm": 0.03206072375178337,
+      "learning_rate": 8.513770180436848e-06,
+      "loss": 0.6338,
+      "step": 5470
+    },
+    {
+      "epoch": 23.004188034188033,
+      "grad_norm": 55.12649154663086,
+      "learning_rate": 8.509021842355176e-06,
+      "loss": 0.9267,
+      "step": 5480
+    },
+    {
+      "epoch": 23.004615384615384,
+      "grad_norm": 68.18521881103516,
+      "learning_rate": 8.504273504273505e-06,
+      "loss": 0.8771,
+      "step": 5490
+    },
+    {
+      "epoch": 23.005042735042736,
+      "grad_norm": 28.95956039428711,
+      "learning_rate": 8.499525166191834e-06,
+      "loss": 0.8377,
+      "step": 5500
+    },
+    {
+      "epoch": 23.005470085470087,
+      "grad_norm": 0.004457338713109493,
+      "learning_rate": 8.494776828110161e-06,
+      "loss": 0.0553,
+      "step": 5510
+    },
+    {
+      "epoch": 23.005897435897435,
+      "grad_norm": 0.024025579914450645,
+      "learning_rate": 8.49002849002849e-06,
+      "loss": 0.8331,
+      "step": 5520
+    },
+    {
+      "epoch": 23.006324786324786,
+      "grad_norm": 0.019380344077944756,
+      "learning_rate": 8.48528015194682e-06,
+      "loss": 1.7575,
+      "step": 5530
+    },
+    {
+      "epoch": 23.006752136752137,
+      "grad_norm": 0.015158622525632381,
+      "learning_rate": 8.480531813865149e-06,
+      "loss": 1.1634,
+      "step": 5540
+    },
+    {
+      "epoch": 23.00717948717949,
+      "grad_norm": 27.653972625732422,
+      "learning_rate": 8.475783475783476e-06,
+      "loss": 1.3936,
+      "step": 5550
+    },
+    {
+      "epoch": 23.007606837606836,
+      "grad_norm": 197.41244506835938,
+      "learning_rate": 8.471035137701806e-06,
+      "loss": 0.9821,
+      "step": 5560
+    },
+    {
+      "epoch": 23.008034188034188,
+      "grad_norm": 0.009685487486422062,
+      "learning_rate": 8.466286799620133e-06,
+      "loss": 0.048,
+      "step": 5570
+    },
+    {
+      "epoch": 23.00846153846154,
+      "grad_norm": 0.04525621980428696,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 1.1907,
+      "step": 5580
+    },
+    {
+      "epoch": 23.00888888888889,
+      "grad_norm": 0.29436931014060974,
+      "learning_rate": 8.456790123456791e-06,
+      "loss": 2.1125,
+      "step": 5590
+    },
+    {
+      "epoch": 23.009316239316238,
+      "grad_norm": 0.028307698667049408,
+      "learning_rate": 8.452041785375119e-06,
+      "loss": 0.6447,
+      "step": 5600
+    },
+    {
+      "epoch": 23.00974358974359,
+      "grad_norm": 171.65646362304688,
+      "learning_rate": 8.447293447293448e-06,
+      "loss": 0.8939,
+      "step": 5610
+    },
+    {
+      "epoch": 23.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 3.0712430477142334,
+      "eval_runtime": 33.4249,
+      "eval_samples_per_second": 0.748,
+      "eval_steps_per_second": 0.748,
+      "step": 5616
+    },
+    {
+      "epoch": 24.00017094017094,
+      "grad_norm": 0.5712462067604065,
+      "learning_rate": 8.442545109211775e-06,
+      "loss": 0.2082,
+      "step": 5620
+    },
+    {
+      "epoch": 24.00059829059829,
+      "grad_norm": 0.008292453363537788,
+      "learning_rate": 8.437796771130106e-06,
+      "loss": 0.6958,
+      "step": 5630
+    },
+    {
+      "epoch": 24.00102564102564,
+      "grad_norm": 118.00381469726562,
+      "learning_rate": 8.433048433048434e-06,
+      "loss": 0.9527,
+      "step": 5640
+    },
+    {
+      "epoch": 24.001452991452993,
+      "grad_norm": 1.618455171585083,
+      "learning_rate": 8.428300094966763e-06,
+      "loss": 0.3995,
+      "step": 5650
+    },
+    {
+      "epoch": 24.00188034188034,
+      "grad_norm": 131.17752075195312,
+      "learning_rate": 8.423551756885092e-06,
+      "loss": 0.3038,
+      "step": 5660
+    },
+    {
+      "epoch": 24.002307692307692,
+      "grad_norm": 2.7916219234466553,
+      "learning_rate": 8.41880341880342e-06,
+      "loss": 0.7865,
+      "step": 5670
+    },
+    {
+      "epoch": 24.002735042735043,
+      "grad_norm": 37.944374084472656,
+      "learning_rate": 8.414055080721748e-06,
+      "loss": 0.1766,
+      "step": 5680
+    },
+    {
+      "epoch": 24.003162393162395,
+      "grad_norm": 0.038069456815719604,
+      "learning_rate": 8.409306742640076e-06,
+      "loss": 0.3233,
+      "step": 5690
+    },
+    {
+      "epoch": 24.003589743589743,
+      "grad_norm": 0.01322519313544035,
+      "learning_rate": 8.404558404558405e-06,
+      "loss": 0.2382,
+      "step": 5700
+    },
+    {
+      "epoch": 24.004017094017094,
+      "grad_norm": 181.4189910888672,
+      "learning_rate": 8.399810066476734e-06,
+      "loss": 0.9649,
+      "step": 5710
+    },
+    {
+      "epoch": 24.004444444444445,
+      "grad_norm": 37.26016616821289,
+      "learning_rate": 8.395061728395062e-06,
+      "loss": 1.6616,
+      "step": 5720
+    },
+    {
+      "epoch": 24.004871794871796,
+      "grad_norm": 0.007100946735590696,
+      "learning_rate": 8.39031339031339e-06,
+      "loss": 1.193,
+      "step": 5730
+    },
+    {
+      "epoch": 24.005299145299144,
+      "grad_norm": 2.50199031829834,
+      "learning_rate": 8.38556505223172e-06,
+      "loss": 1.475,
+      "step": 5740
+    },
+    {
+      "epoch": 24.005726495726496,
+      "grad_norm": 0.008331292308866978,
+      "learning_rate": 8.380816714150049e-06,
+      "loss": 0.2329,
+      "step": 5750
+    },
+    {
+      "epoch": 24.006153846153847,
+      "grad_norm": 0.04707374423742294,
+      "learning_rate": 8.376068376068377e-06,
+      "loss": 0.363,
+      "step": 5760
+    },
+    {
+      "epoch": 24.006581196581198,
+      "grad_norm": 0.08683554083108902,
+      "learning_rate": 8.371320037986706e-06,
+      "loss": 1.9591,
+      "step": 5770
+    },
+    {
+      "epoch": 24.007008547008546,
+      "grad_norm": 0.020473746582865715,
+      "learning_rate": 8.366571699905033e-06,
+      "loss": 1.562,
+      "step": 5780
+    },
+    {
+      "epoch": 24.007435897435897,
+      "grad_norm": 48.53575134277344,
+      "learning_rate": 8.361823361823362e-06,
+      "loss": 0.5132,
+      "step": 5790
+    },
+    {
+      "epoch": 24.00786324786325,
+      "grad_norm": 0.007784120738506317,
+      "learning_rate": 8.357075023741691e-06,
+      "loss": 1.1691,
+      "step": 5800
+    },
+    {
+      "epoch": 24.0082905982906,
+      "grad_norm": 0.08522366732358932,
+      "learning_rate": 8.352326685660019e-06,
+      "loss": 0.0145,
+      "step": 5810
+    },
+    {
+      "epoch": 24.008717948717948,
+      "grad_norm": 3.588449001312256,
+      "learning_rate": 8.347578347578348e-06,
+      "loss": 0.2557,
+      "step": 5820
+    },
+    {
+      "epoch": 24.0091452991453,
+      "grad_norm": 6.431326389312744,
+      "learning_rate": 8.342830009496677e-06,
+      "loss": 0.8734,
+      "step": 5830
+    },
+    {
+      "epoch": 24.00957264957265,
+      "grad_norm": 10.404790878295898,
+      "learning_rate": 8.338081671415006e-06,
+      "loss": 1.3251,
+      "step": 5840
+    },
+    {
+      "epoch": 24.01,
+      "grad_norm": 0.02091125212609768,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.3871,
+      "step": 5850
+    },
+    {
+      "epoch": 24.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 3.2393715381622314,
+      "eval_runtime": 33.9657,
+      "eval_samples_per_second": 0.736,
+      "eval_steps_per_second": 0.736,
+      "step": 5850
+    },
+    {
+      "epoch": 25.00042735042735,
+      "grad_norm": 134.0677947998047,
+      "learning_rate": 8.328584995251663e-06,
+      "loss": 0.6986,
+      "step": 5860
+    },
+    {
+      "epoch": 25.000854700854703,
+      "grad_norm": 0.03229491412639618,
+      "learning_rate": 8.323836657169992e-06,
+      "loss": 0.768,
+      "step": 5870
+    },
+    {
+      "epoch": 25.00128205128205,
+      "grad_norm": 156.05152893066406,
+      "learning_rate": 8.31908831908832e-06,
+      "loss": 1.0463,
+      "step": 5880
+    },
+    {
+      "epoch": 25.0017094017094,
+      "grad_norm": 0.014203476719558239,
+      "learning_rate": 8.314339981006649e-06,
+      "loss": 0.488,
+      "step": 5890
+    },
+    {
+      "epoch": 25.002136752136753,
+      "grad_norm": 285.82000732421875,
+      "learning_rate": 8.309591642924976e-06,
+      "loss": 0.8084,
+      "step": 5900
+    },
+    {
+      "epoch": 25.002564102564104,
+      "grad_norm": 122.90914154052734,
+      "learning_rate": 8.304843304843305e-06,
+      "loss": 0.8786,
+      "step": 5910
+    },
+    {
+      "epoch": 25.002991452991452,
+      "grad_norm": 8.224214553833008,
+      "learning_rate": 8.300094966761634e-06,
+      "loss": 0.3977,
+      "step": 5920
+    },
+    {
+      "epoch": 25.003418803418803,
+      "grad_norm": 230.34385681152344,
+      "learning_rate": 8.295346628679962e-06,
+      "loss": 0.5285,
+      "step": 5930
+    },
+    {
+      "epoch": 25.003846153846155,
+      "grad_norm": 1.262283205986023,
+      "learning_rate": 8.290598290598293e-06,
+      "loss": 0.8955,
+      "step": 5940
+    },
+    {
+      "epoch": 25.004273504273506,
+      "grad_norm": 71.28157043457031,
+      "learning_rate": 8.28584995251662e-06,
+      "loss": 0.53,
+      "step": 5950
+    },
+    {
+      "epoch": 25.004700854700854,
+      "grad_norm": 0.19136950373649597,
+      "learning_rate": 8.28110161443495e-06,
+      "loss": 0.8647,
+      "step": 5960
+    },
+    {
+      "epoch": 25.005128205128205,
+      "grad_norm": 0.2358364760875702,
+      "learning_rate": 8.276353276353277e-06,
+      "loss": 0.0021,
+      "step": 5970
+    },
+    {
+      "epoch": 25.005555555555556,
+      "grad_norm": 0.007490760181099176,
+      "learning_rate": 8.271604938271606e-06,
+      "loss": 0.6602,
+      "step": 5980
+    },
+    {
+      "epoch": 25.005982905982908,
+      "grad_norm": 0.006508139427751303,
+      "learning_rate": 8.266856600189935e-06,
+      "loss": 0.7927,
+      "step": 5990
+    },
+    {
+      "epoch": 25.006410256410255,
+      "grad_norm": 0.024243731051683426,
+      "learning_rate": 8.262108262108262e-06,
+      "loss": 1.5309,
+      "step": 6000
+    },
+    {
+      "epoch": 25.006837606837607,
+      "grad_norm": 11.39758586883545,
+      "learning_rate": 8.257359924026592e-06,
+      "loss": 0.311,
+      "step": 6010
+    },
+    {
+      "epoch": 25.007264957264958,
+      "grad_norm": 0.23802872002124786,
+      "learning_rate": 8.252611585944919e-06,
+      "loss": 0.0891,
+      "step": 6020
+    },
+    {
+      "epoch": 25.00769230769231,
+      "grad_norm": 14.19528865814209,
+      "learning_rate": 8.247863247863248e-06,
+      "loss": 1.7728,
+      "step": 6030
+    },
+    {
+      "epoch": 25.008119658119657,
+      "grad_norm": 0.0041798874735832214,
+      "learning_rate": 8.243114909781577e-06,
+      "loss": 0.0076,
+      "step": 6040
+    },
+    {
+      "epoch": 25.00854700854701,
+      "grad_norm": 83.91490936279297,
+      "learning_rate": 8.238366571699906e-06,
+      "loss": 1.0939,
+      "step": 6050
+    },
+    {
+      "epoch": 25.00897435897436,
+      "grad_norm": 109.98046875,
+      "learning_rate": 8.233618233618234e-06,
+      "loss": 0.7906,
+      "step": 6060
+    },
+    {
+      "epoch": 25.00940170940171,
+      "grad_norm": 0.0034200188238173723,
+      "learning_rate": 8.228869895536563e-06,
+      "loss": 0.3631,
+      "step": 6070
+    },
+    {
+      "epoch": 25.00982905982906,
+      "grad_norm": 0.009681742638349533,
+      "learning_rate": 8.224121557454892e-06,
+      "loss": 0.3649,
+      "step": 6080
+    },
+    {
+      "epoch": 25.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 3.9465994834899902,
+      "eval_runtime": 33.084,
+      "eval_samples_per_second": 0.756,
+      "eval_steps_per_second": 0.756,
+      "step": 6084
+    },
+    {
+      "epoch": 26.00025641025641,
+      "grad_norm": 0.19076739251613617,
+      "learning_rate": 8.21937321937322e-06,
+      "loss": 0.9254,
+      "step": 6090
+    },
+    {
+      "epoch": 26.00068376068376,
+      "grad_norm": 0.0032322730403393507,
+      "learning_rate": 8.214624881291549e-06,
+      "loss": 0.7628,
+      "step": 6100
+    },
+    {
+      "epoch": 26.00111111111111,
+      "grad_norm": 78.26629638671875,
+      "learning_rate": 8.209876543209876e-06,
+      "loss": 1.0775,
+      "step": 6110
+    },
+    {
+      "epoch": 26.001538461538463,
+      "grad_norm": 41.407955169677734,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 0.6687,
+      "step": 6120
+    },
+    {
+      "epoch": 26.00196581196581,
+      "grad_norm": 0.018887581303715706,
+      "learning_rate": 8.200379867046535e-06,
+      "loss": 0.6125,
+      "step": 6130
+    },
+    {
+      "epoch": 26.00239316239316,
+      "grad_norm": 0.004367130342870951,
+      "learning_rate": 8.195631528964862e-06,
+      "loss": 0.4048,
+      "step": 6140
+    },
+    {
+      "epoch": 26.002820512820513,
+      "grad_norm": 0.005417748354375362,
+      "learning_rate": 8.190883190883193e-06,
+      "loss": 0.1982,
+      "step": 6150
+    },
+    {
+      "epoch": 26.003247863247864,
+      "grad_norm": 0.028607884421944618,
+      "learning_rate": 8.18613485280152e-06,
+      "loss": 1.1038,
+      "step": 6160
+    },
+    {
+      "epoch": 26.003675213675212,
+      "grad_norm": 167.5290069580078,
+      "learning_rate": 8.18138651471985e-06,
+      "loss": 0.2927,
+      "step": 6170
+    },
+    {
+      "epoch": 26.004102564102563,
+      "grad_norm": 0.5390312671661377,
+      "learning_rate": 8.176638176638177e-06,
+      "loss": 1.0214,
+      "step": 6180
+    },
+    {
+      "epoch": 26.004529914529915,
+      "grad_norm": 0.007749592885375023,
+      "learning_rate": 8.171889838556506e-06,
+      "loss": 0.2385,
+      "step": 6190
+    },
+    {
+      "epoch": 26.004957264957266,
+      "grad_norm": 0.005856471136212349,
+      "learning_rate": 8.167141500474835e-06,
+      "loss": 0.2423,
+      "step": 6200
+    },
+    {
+      "epoch": 26.005384615384614,
+      "grad_norm": 136.56642150878906,
+      "learning_rate": 8.162393162393163e-06,
+      "loss": 0.7251,
+      "step": 6210
+    },
+    {
+      "epoch": 26.005811965811965,
+      "grad_norm": 482.7048034667969,
+      "learning_rate": 8.157644824311492e-06,
+      "loss": 0.7191,
+      "step": 6220
+    },
+    {
+      "epoch": 26.006239316239316,
+      "grad_norm": 0.00340673909522593,
+      "learning_rate": 8.15289648622982e-06,
+      "loss": 0.0008,
+      "step": 6230
+    },
+    {
+      "epoch": 26.006666666666668,
+      "grad_norm": 0.0029049227014184,
+      "learning_rate": 8.148148148148148e-06,
+      "loss": 0.7355,
+      "step": 6240
+    },
+    {
+      "epoch": 26.007094017094015,
+      "grad_norm": 0.15764747560024261,
+      "learning_rate": 8.143399810066477e-06,
+      "loss": 0.8012,
+      "step": 6250
+    },
+    {
+      "epoch": 26.007521367521367,
+      "grad_norm": 0.00497576454654336,
+      "learning_rate": 8.138651471984807e-06,
+      "loss": 0.7658,
+      "step": 6260
+    },
+    {
+      "epoch": 26.007948717948718,
+      "grad_norm": 0.009462869726121426,
+      "learning_rate": 8.133903133903136e-06,
+      "loss": 0.001,
+      "step": 6270
+    },
+    {
+      "epoch": 26.00837606837607,
+      "grad_norm": 0.006881650071591139,
+      "learning_rate": 8.129154795821463e-06,
+      "loss": 0.0038,
+      "step": 6280
+    },
+    {
+      "epoch": 26.008803418803417,
+      "grad_norm": 0.036327850073575974,
+      "learning_rate": 8.124406457739792e-06,
+      "loss": 3.0791,
+      "step": 6290
+    },
+    {
+      "epoch": 26.00923076923077,
+      "grad_norm": 78.0274658203125,
+      "learning_rate": 8.11965811965812e-06,
+      "loss": 0.3814,
+      "step": 6300
+    },
+    {
+      "epoch": 26.00965811965812,
+      "grad_norm": 0.21120551228523254,
+      "learning_rate": 8.114909781576449e-06,
+      "loss": 1.2601,
+      "step": 6310
+    },
+    {
+      "epoch": 26.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 2.9586069583892822,
+      "eval_runtime": 33.5582,
+      "eval_samples_per_second": 0.745,
+      "eval_steps_per_second": 0.745,
+      "step": 6318
+    },
+    {
+      "epoch": 27.00008547008547,
+      "grad_norm": 67.51969146728516,
+      "learning_rate": 8.110161443494778e-06,
+      "loss": 0.9303,
+      "step": 6320
+    },
+    {
+      "epoch": 27.00051282051282,
+      "grad_norm": 104.40100860595703,
+      "learning_rate": 8.105413105413106e-06,
+      "loss": 0.4351,
+      "step": 6330
+    },
+    {
+      "epoch": 27.000940170940172,
+      "grad_norm": 0.016888683661818504,
+      "learning_rate": 8.100664767331435e-06,
+      "loss": 0.3423,
+      "step": 6340
+    },
+    {
+      "epoch": 27.00136752136752,
+      "grad_norm": 0.011165400967001915,
+      "learning_rate": 8.095916429249762e-06,
+      "loss": 1.4128,
+      "step": 6350
+    },
+    {
+      "epoch": 27.00179487179487,
+      "grad_norm": 0.027417538687586784,
+      "learning_rate": 8.091168091168093e-06,
+      "loss": 0.002,
+      "step": 6360
+    },
+    {
+      "epoch": 27.002222222222223,
+      "grad_norm": 0.09121936559677124,
+      "learning_rate": 8.08641975308642e-06,
+      "loss": 2.5068,
+      "step": 6370
+    },
+    {
+      "epoch": 27.002649572649574,
+      "grad_norm": 0.007114642299711704,
+      "learning_rate": 8.08167141500475e-06,
+      "loss": 0.1004,
+      "step": 6380
+    },
+    {
+      "epoch": 27.00307692307692,
+      "grad_norm": 0.03234013542532921,
+      "learning_rate": 8.076923076923077e-06,
+      "loss": 0.2587,
+      "step": 6390
+    },
+    {
+      "epoch": 27.003504273504273,
+      "grad_norm": 0.06352563947439194,
+      "learning_rate": 8.072174738841406e-06,
+      "loss": 0.2638,
+      "step": 6400
+    },
+    {
+      "epoch": 27.003931623931624,
+      "grad_norm": 0.03514587879180908,
+      "learning_rate": 8.067426400759735e-06,
+      "loss": 0.0102,
+      "step": 6410
+    },
+    {
+      "epoch": 27.004358974358976,
+      "grad_norm": 0.08466164022684097,
+      "learning_rate": 8.062678062678063e-06,
+      "loss": 0.3654,
+      "step": 6420
+    },
+    {
+      "epoch": 27.004786324786323,
+      "grad_norm": 0.01347273774445057,
+      "learning_rate": 8.057929724596392e-06,
+      "loss": 0.0482,
+      "step": 6430
+    },
+    {
+      "epoch": 27.005213675213675,
+      "grad_norm": 25.30817985534668,
+      "learning_rate": 8.05318138651472e-06,
+      "loss": 0.0527,
+      "step": 6440
+    },
+    {
+      "epoch": 27.005641025641026,
+      "grad_norm": 0.003483937354758382,
+      "learning_rate": 8.048433048433048e-06,
+      "loss": 0.5832,
+      "step": 6450
+    },
+    {
+      "epoch": 27.006068376068377,
+      "grad_norm": 0.07294996827840805,
+      "learning_rate": 8.043684710351378e-06,
+      "loss": 0.5054,
+      "step": 6460
+    },
+    {
+      "epoch": 27.006495726495725,
+      "grad_norm": 0.02032245323061943,
+      "learning_rate": 8.038936372269707e-06,
+      "loss": 0.2399,
+      "step": 6470
+    },
+    {
+      "epoch": 27.006923076923076,
+      "grad_norm": 5.702065944671631,
+      "learning_rate": 8.034188034188036e-06,
+      "loss": 1.0618,
+      "step": 6480
+    },
+    {
+      "epoch": 27.007350427350428,
+      "grad_norm": 0.2200225442647934,
+      "learning_rate": 8.029439696106363e-06,
+      "loss": 0.4855,
+      "step": 6490
+    },
+    {
+      "epoch": 27.00777777777778,
+      "grad_norm": 0.02679477632045746,
+      "learning_rate": 8.024691358024692e-06,
+      "loss": 0.9622,
+      "step": 6500
+    },
+    {
+      "epoch": 27.008205128205127,
+      "grad_norm": 0.00978939514607191,
+      "learning_rate": 8.01994301994302e-06,
+      "loss": 0.173,
+      "step": 6510
+    },
+    {
+      "epoch": 27.008632478632478,
+      "grad_norm": 0.009447705931961536,
+      "learning_rate": 8.015194681861349e-06,
+      "loss": 0.0833,
+      "step": 6520
+    },
+    {
+      "epoch": 27.00905982905983,
+      "grad_norm": 6.86177921295166,
+      "learning_rate": 8.010446343779678e-06,
+      "loss": 0.3262,
+      "step": 6530
+    },
+    {
+      "epoch": 27.00948717948718,
+      "grad_norm": 0.2438230961561203,
+      "learning_rate": 8.005698005698006e-06,
+      "loss": 0.1945,
+      "step": 6540
+    },
+    {
+      "epoch": 27.00991452991453,
+      "grad_norm": 0.0071107167750597,
+      "learning_rate": 8.000949667616335e-06,
+      "loss": 0.852,
+      "step": 6550
+    },
+    {
+      "epoch": 27.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 4.646361827850342,
+      "eval_runtime": 33.5078,
+      "eval_samples_per_second": 0.746,
+      "eval_steps_per_second": 0.746,
+      "step": 6552
+    },
+    {
+      "epoch": 28.00034188034188,
+      "grad_norm": 0.007593700196594,
+      "learning_rate": 7.996201329534662e-06,
+      "loss": 0.9613,
+      "step": 6560
+    },
+    {
+      "epoch": 28.00076923076923,
+      "grad_norm": 198.82052612304688,
+      "learning_rate": 7.991452991452993e-06,
+      "loss": 0.6948,
+      "step": 6570
+    },
+    {
+      "epoch": 28.00119658119658,
+      "grad_norm": 0.006188979372382164,
+      "learning_rate": 7.98670465337132e-06,
+      "loss": 1.0506,
+      "step": 6580
+    },
+    {
+      "epoch": 28.001623931623932,
+      "grad_norm": 0.0029619692359119654,
+      "learning_rate": 7.98195631528965e-06,
+      "loss": 0.5341,
+      "step": 6590
+    },
+    {
+      "epoch": 28.002051282051283,
+      "grad_norm": 0.017630403861403465,
+      "learning_rate": 7.977207977207979e-06,
+      "loss": 0.775,
+      "step": 6600
+    },
+    {
+      "epoch": 28.00247863247863,
+      "grad_norm": 0.15129226446151733,
+      "learning_rate": 7.972459639126306e-06,
+      "loss": 0.3747,
+      "step": 6610
+    },
+    {
+      "epoch": 28.002905982905983,
+      "grad_norm": 0.11231208592653275,
+      "learning_rate": 7.967711301044635e-06,
+      "loss": 0.0478,
+      "step": 6620
+    },
+    {
+      "epoch": 28.003333333333334,
+      "grad_norm": 0.01004121731966734,
+      "learning_rate": 7.962962962962963e-06,
+      "loss": 0.5073,
+      "step": 6630
+    },
+    {
+      "epoch": 28.003760683760685,
+      "grad_norm": 0.0024136013817042112,
+      "learning_rate": 7.958214624881292e-06,
+      "loss": 1.5819,
+      "step": 6640
+    },
+    {
+      "epoch": 28.004188034188033,
+      "grad_norm": 329.79754638671875,
+      "learning_rate": 7.953466286799621e-06,
+      "loss": 0.5354,
+      "step": 6650
+    },
+    {
+      "epoch": 28.004615384615384,
+      "grad_norm": 9.606019973754883,
+      "learning_rate": 7.948717948717949e-06,
+      "loss": 0.0239,
+      "step": 6660
+    },
+    {
+      "epoch": 28.005042735042736,
+      "grad_norm": 161.5113067626953,
+      "learning_rate": 7.943969610636278e-06,
+      "loss": 1.0728,
+      "step": 6670
+    },
+    {
+      "epoch": 28.005470085470087,
+      "grad_norm": 0.014358972199261189,
+      "learning_rate": 7.939221272554607e-06,
+      "loss": 0.0171,
+      "step": 6680
+    },
+    {
+      "epoch": 28.005897435897435,
+      "grad_norm": 96.80726623535156,
+      "learning_rate": 7.934472934472936e-06,
+      "loss": 0.5362,
+      "step": 6690
+    },
+    {
+      "epoch": 28.006324786324786,
+      "grad_norm": 0.0033912707585841417,
+      "learning_rate": 7.929724596391264e-06,
+      "loss": 0.6892,
+      "step": 6700
+    },
+    {
+      "epoch": 28.006752136752137,
+      "grad_norm": 11.932380676269531,
+      "learning_rate": 7.924976258309593e-06,
+      "loss": 0.4339,
+      "step": 6710
+    },
+    {
+      "epoch": 28.00717948717949,
+      "grad_norm": 0.026500051841139793,
+      "learning_rate": 7.92022792022792e-06,
+      "loss": 1.4133,
+      "step": 6720
+    },
+    {
+      "epoch": 28.007606837606836,
+      "grad_norm": 0.003824119921773672,
+      "learning_rate": 7.91547958214625e-06,
+      "loss": 0.8,
+      "step": 6730
+    },
+    {
+      "epoch": 28.008034188034188,
+      "grad_norm": 0.02020156756043434,
+      "learning_rate": 7.910731244064578e-06,
+      "loss": 0.9052,
+      "step": 6740
+    },
+    {
+      "epoch": 28.00846153846154,
+      "grad_norm": 80.97819519042969,
+      "learning_rate": 7.905982905982906e-06,
+      "loss": 0.2557,
+      "step": 6750
+    },
+    {
+      "epoch": 28.00888888888889,
+      "grad_norm": 79.75435638427734,
+      "learning_rate": 7.901234567901235e-06,
+      "loss": 0.7378,
+      "step": 6760
+    },
+    {
+      "epoch": 28.009316239316238,
+      "grad_norm": 0.6139788031578064,
+      "learning_rate": 7.896486229819562e-06,
+      "loss": 0.595,
+      "step": 6770
+    },
+    {
+      "epoch": 28.00974358974359,
+      "grad_norm": 181.7918243408203,
+      "learning_rate": 7.891737891737893e-06,
+      "loss": 0.6269,
+      "step": 6780
+    },
+    {
+      "epoch": 28.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 3.1291773319244385,
+      "eval_runtime": 33.0523,
+      "eval_samples_per_second": 0.756,
+      "eval_steps_per_second": 0.756,
+      "step": 6786
+    },
+    {
+      "epoch": 29.00017094017094,
+      "grad_norm": 0.010640930384397507,
+      "learning_rate": 7.88698955365622e-06,
+      "loss": 0.8873,
+      "step": 6790
+    },
+    {
+      "epoch": 29.00059829059829,
+      "grad_norm": 0.007870799861848354,
+      "learning_rate": 7.88224121557455e-06,
+      "loss": 0.013,
+      "step": 6800
+    },
+    {
+      "epoch": 29.00102564102564,
+      "grad_norm": 428.0439453125,
+      "learning_rate": 7.877492877492879e-06,
+      "loss": 0.3937,
+      "step": 6810
+    },
+    {
+      "epoch": 29.001452991452993,
+      "grad_norm": 0.003777164965867996,
+      "learning_rate": 7.872744539411206e-06,
+      "loss": 0.2548,
+      "step": 6820
+    },
+    {
+      "epoch": 29.00188034188034,
+      "grad_norm": 754.8897094726562,
+      "learning_rate": 7.867996201329536e-06,
+      "loss": 0.7434,
+      "step": 6830
+    },
+    {
+      "epoch": 29.002307692307692,
+      "grad_norm": 0.002904111985117197,
+      "learning_rate": 7.863247863247863e-06,
+      "loss": 0.6923,
+      "step": 6840
+    },
+    {
+      "epoch": 29.002735042735043,
+      "grad_norm": 405.4997253417969,
+      "learning_rate": 7.858499525166192e-06,
+      "loss": 1.4677,
+      "step": 6850
+    },
+    {
+      "epoch": 29.003162393162395,
+      "grad_norm": 0.03453909978270531,
+      "learning_rate": 7.853751187084521e-06,
+      "loss": 0.2134,
+      "step": 6860
+    },
+    {
+      "epoch": 29.003589743589743,
+      "grad_norm": 0.002487962134182453,
+      "learning_rate": 7.849002849002849e-06,
+      "loss": 0.0651,
+      "step": 6870
+    },
+    {
+      "epoch": 29.004017094017094,
+      "grad_norm": 0.0023798488546162844,
+      "learning_rate": 7.844254510921178e-06,
+      "loss": 0.8198,
+      "step": 6880
+    },
+    {
+      "epoch": 29.004444444444445,
+      "grad_norm": 0.0023906242568045855,
+      "learning_rate": 7.839506172839507e-06,
+      "loss": 1.3586,
+      "step": 6890
+    },
+    {
+      "epoch": 29.004871794871796,
+      "grad_norm": 0.0033716100733727217,
+      "learning_rate": 7.834757834757836e-06,
+      "loss": 0.0181,
+      "step": 6900
+    },
+    {
+      "epoch": 29.005299145299144,
+      "grad_norm": 260.02825927734375,
+      "learning_rate": 7.830009496676164e-06,
+      "loss": 0.287,
+      "step": 6910
+    },
+    {
+      "epoch": 29.005726495726496,
+      "grad_norm": 0.00520634651184082,
+      "learning_rate": 7.825261158594493e-06,
+      "loss": 0.0378,
+      "step": 6920
+    },
+    {
+      "epoch": 29.006153846153847,
+      "grad_norm": 0.005019763018935919,
+      "learning_rate": 7.820512820512822e-06,
+      "loss": 1.2482,
+      "step": 6930
+    },
+    {
+      "epoch": 29.006581196581198,
+      "grad_norm": 136.67764282226562,
+      "learning_rate": 7.81576448243115e-06,
+      "loss": 1.2099,
+      "step": 6940
+    },
+    {
+      "epoch": 29.007008547008546,
+      "grad_norm": 3.6764883995056152,
+      "learning_rate": 7.811016144349479e-06,
+      "loss": 1.5087,
+      "step": 6950
+    },
+    {
+      "epoch": 29.007435897435897,
+      "grad_norm": 0.004404416773468256,
+      "learning_rate": 7.806267806267806e-06,
+      "loss": 0.5894,
+      "step": 6960
+    },
+    {
+      "epoch": 29.00786324786325,
+      "grad_norm": 0.008280979469418526,
+      "learning_rate": 7.801519468186135e-06,
+      "loss": 0.4367,
+      "step": 6970
+    },
+    {
+      "epoch": 29.0082905982906,
+      "grad_norm": 0.006669474299997091,
+      "learning_rate": 7.796771130104464e-06,
+      "loss": 0.5476,
+      "step": 6980
+    },
+    {
+      "epoch": 29.008717948717948,
+      "grad_norm": 0.012090092524886131,
+      "learning_rate": 7.792022792022793e-06,
+      "loss": 1.0435,
+      "step": 6990
+    },
+    {
+      "epoch": 29.0091452991453,
+      "grad_norm": 0.0026431807782500982,
+      "learning_rate": 7.787274453941121e-06,
+      "loss": 0.0031,
+      "step": 7000
+    },
+    {
+      "epoch": 29.00957264957265,
+      "grad_norm": 0.012273382395505905,
+      "learning_rate": 7.78252611585945e-06,
+      "loss": 0.7421,
+      "step": 7010
+    },
+    {
+      "epoch": 29.01,
+      "grad_norm": 0.3034244179725647,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 1.0013,
+      "step": 7020
+    },
+    {
+      "epoch": 29.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 4.631889820098877,
+      "eval_runtime": 33.5346,
+      "eval_samples_per_second": 0.745,
+      "eval_steps_per_second": 0.745,
+      "step": 7020
+    },
+    {
+      "epoch": 30.00042735042735,
+      "grad_norm": 1.9434560537338257,
+      "learning_rate": 7.773029439696107e-06,
+      "loss": 0.8706,
+      "step": 7030
+    },
+    {
+      "epoch": 30.000854700854703,
+      "grad_norm": 0.003200115170329809,
+      "learning_rate": 7.768281101614436e-06,
+      "loss": 0.5544,
+      "step": 7040
+    },
+    {
+      "epoch": 30.00128205128205,
+      "grad_norm": 0.005437285173684359,
+      "learning_rate": 7.763532763532763e-06,
+      "loss": 1.3336,
+      "step": 7050
+    },
+    {
+      "epoch": 30.0017094017094,
+      "grad_norm": 0.4712231457233429,
+      "learning_rate": 7.758784425451092e-06,
+      "loss": 1.0848,
+      "step": 7060
+    },
+    {
+      "epoch": 30.002136752136753,
+      "grad_norm": 0.0037900730967521667,
+      "learning_rate": 7.754036087369421e-06,
+      "loss": 0.3756,
+      "step": 7070
+    },
+    {
+      "epoch": 30.002564102564104,
+      "grad_norm": 114.87593841552734,
+      "learning_rate": 7.749287749287749e-06,
+      "loss": 0.6611,
+      "step": 7080
+    },
+    {
+      "epoch": 30.002991452991452,
+      "grad_norm": 0.08566898107528687,
+      "learning_rate": 7.74453941120608e-06,
+      "loss": 0.2429,
+      "step": 7090
+    },
+    {
+      "epoch": 30.003418803418803,
+      "grad_norm": 0.0038280077278614044,
+      "learning_rate": 7.739791073124407e-06,
+      "loss": 0.001,
+      "step": 7100
+    },
+    {
+      "epoch": 30.003846153846155,
+      "grad_norm": 0.008761651813983917,
+      "learning_rate": 7.735042735042736e-06,
+      "loss": 0.0407,
+      "step": 7110
+    },
+    {
+      "epoch": 30.004273504273506,
+      "grad_norm": 0.0022454620338976383,
+      "learning_rate": 7.730294396961064e-06,
+      "loss": 0.8858,
+      "step": 7120
+    },
+    {
+      "epoch": 30.004700854700854,
+      "grad_norm": 0.007320001721382141,
+      "learning_rate": 7.725546058879393e-06,
+      "loss": 0.5604,
+      "step": 7130
+    },
+    {
+      "epoch": 30.005128205128205,
+      "grad_norm": 99.08616638183594,
+      "learning_rate": 7.720797720797722e-06,
+      "loss": 0.727,
+      "step": 7140
+    },
+    {
+      "epoch": 30.005555555555556,
+      "grad_norm": 0.013165280222892761,
+      "learning_rate": 7.71604938271605e-06,
+      "loss": 0.0583,
+      "step": 7150
+    },
+    {
+      "epoch": 30.005982905982908,
+      "grad_norm": 0.0017615576507523656,
+      "learning_rate": 7.711301044634379e-06,
+      "loss": 0.435,
+      "step": 7160
+    },
+    {
+      "epoch": 30.006410256410255,
+      "grad_norm": 0.008409475907683372,
+      "learning_rate": 7.706552706552706e-06,
+      "loss": 0.7466,
+      "step": 7170
+    },
+    {
+      "epoch": 30.006837606837607,
+      "grad_norm": 0.002216171706095338,
+      "learning_rate": 7.701804368471035e-06,
+      "loss": 0.6197,
+      "step": 7180
+    },
+    {
+      "epoch": 30.007264957264958,
+      "grad_norm": 0.015576567500829697,
+      "learning_rate": 7.697056030389364e-06,
+      "loss": 0.0003,
+      "step": 7190
+    },
+    {
+      "epoch": 30.00769230769231,
+      "grad_norm": 0.002197784371674061,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 0.0428,
+      "step": 7200
+    },
+    {
+      "epoch": 30.008119658119657,
+      "grad_norm": 0.019170358777046204,
+      "learning_rate": 7.687559354226021e-06,
+      "loss": 2.7654,
+      "step": 7210
+    },
+    {
+      "epoch": 30.00854700854701,
+      "grad_norm": 0.0038994126953184605,
+      "learning_rate": 7.68281101614435e-06,
+      "loss": 1.1708,
+      "step": 7220
+    },
+    {
+      "epoch": 30.00897435897436,
+      "grad_norm": 0.17869791388511658,
+      "learning_rate": 7.67806267806268e-06,
+      "loss": 0.0286,
+      "step": 7230
+    },
+    {
+      "epoch": 30.00940170940171,
+      "grad_norm": 0.006502739619463682,
+      "learning_rate": 7.673314339981007e-06,
+      "loss": 0.4125,
+      "step": 7240
+    },
+    {
+      "epoch": 30.00982905982906,
+      "grad_norm": 0.01781412959098816,
+      "learning_rate": 7.668566001899336e-06,
+      "loss": 0.02,
+      "step": 7250
+    },
+    {
+      "epoch": 30.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 4.251375198364258,
+      "eval_runtime": 33.6135,
+      "eval_samples_per_second": 0.744,
+      "eval_steps_per_second": 0.744,
+      "step": 7254
+    },
+    {
+      "epoch": 31.00025641025641,
+      "grad_norm": 0.3418414294719696,
+      "learning_rate": 7.663817663817665e-06,
+      "loss": 0.4944,
+      "step": 7260
+    },
+    {
+      "epoch": 31.00068376068376,
+      "grad_norm": 0.08820147812366486,
+      "learning_rate": 7.659069325735993e-06,
+      "loss": 0.6409,
+      "step": 7270
+    },
+    {
+      "epoch": 31.00111111111111,
+      "grad_norm": 0.004805733449757099,
+      "learning_rate": 7.654320987654322e-06,
+      "loss": 0.703,
+      "step": 7280
+    },
+    {
+      "epoch": 31.001538461538463,
+      "grad_norm": 171.22964477539062,
+      "learning_rate": 7.649572649572649e-06,
+      "loss": 0.3897,
+      "step": 7290
+    },
+    {
+      "epoch": 31.00196581196581,
+      "grad_norm": 0.0735219419002533,
+      "learning_rate": 7.64482431149098e-06,
+      "loss": 1.2327,
+      "step": 7300
+    },
+    {
+      "epoch": 31.00239316239316,
+      "grad_norm": 0.19371773302555084,
+      "learning_rate": 7.640075973409307e-06,
+      "loss": 0.0612,
+      "step": 7310
+    },
+    {
+      "epoch": 31.002820512820513,
+      "grad_norm": 0.011904860846698284,
+      "learning_rate": 7.635327635327637e-06,
+      "loss": 0.0011,
+      "step": 7320
+    },
+    {
+      "epoch": 31.003247863247864,
+      "grad_norm": 53.154327392578125,
+      "learning_rate": 7.630579297245964e-06,
+      "loss": 0.032,
+      "step": 7330
+    },
+    {
+      "epoch": 31.003675213675212,
+      "grad_norm": 0.030124612152576447,
+      "learning_rate": 7.625830959164293e-06,
+      "loss": 0.171,
+      "step": 7340
+    },
+    {
+      "epoch": 31.004102564102563,
+      "grad_norm": 0.005205671768635511,
+      "learning_rate": 7.6210826210826214e-06,
+      "loss": 0.5295,
+      "step": 7350
+    },
+    {
+      "epoch": 31.004529914529915,
+      "grad_norm": 3.4219090938568115,
+      "learning_rate": 7.61633428300095e-06,
+      "loss": 0.5748,
+      "step": 7360
+    },
+    {
+      "epoch": 31.004957264957266,
+      "grad_norm": 0.027771934866905212,
+      "learning_rate": 7.611585944919279e-06,
+      "loss": 0.0483,
+      "step": 7370
+    },
+    {
+      "epoch": 31.005384615384614,
+      "grad_norm": 0.0065915524028241634,
+      "learning_rate": 7.606837606837607e-06,
+      "loss": 0.8829,
+      "step": 7380
+    },
+    {
+      "epoch": 31.005811965811965,
+      "grad_norm": 0.01646728627383709,
+      "learning_rate": 7.6020892687559355e-06,
+      "loss": 0.0029,
+      "step": 7390
+    },
+    {
+      "epoch": 31.006239316239316,
+      "grad_norm": 0.007259257137775421,
+      "learning_rate": 7.5973409306742655e-06,
+      "loss": 0.8513,
+      "step": 7400
+    },
+    {
+      "epoch": 31.006666666666668,
+      "grad_norm": 0.004463976714760065,
+      "learning_rate": 7.592592592592594e-06,
+      "loss": 0.4636,
+      "step": 7410
+    },
+    {
+      "epoch": 31.007094017094015,
+      "grad_norm": 0.0043134731240570545,
+      "learning_rate": 7.587844254510922e-06,
+      "loss": 1.0482,
+      "step": 7420
+    },
+    {
+      "epoch": 31.007521367521367,
+      "grad_norm": 0.006201723124831915,
+      "learning_rate": 7.58309591642925e-06,
+      "loss": 0.2689,
+      "step": 7430
+    },
+    {
+      "epoch": 31.007948717948718,
+      "grad_norm": 17.315887451171875,
+      "learning_rate": 7.578347578347579e-06,
+      "loss": 1.319,
+      "step": 7440
+    },
+    {
+      "epoch": 31.00837606837607,
+      "grad_norm": 0.005171514581888914,
+      "learning_rate": 7.573599240265908e-06,
+      "loss": 0.002,
+      "step": 7450
+    },
+    {
+      "epoch": 31.008803418803417,
+      "grad_norm": 108.37242889404297,
+      "learning_rate": 7.568850902184236e-06,
+      "loss": 0.9821,
+      "step": 7460
+    },
+    {
+      "epoch": 31.00923076923077,
+      "grad_norm": 3.896742820739746,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 1.0268,
+      "step": 7470
+    },
+    {
+      "epoch": 31.00965811965812,
+      "grad_norm": 0.0028510969132184982,
+      "learning_rate": 7.559354226020893e-06,
+      "loss": 0.1333,
+      "step": 7480
+    },
+    {
+      "epoch": 31.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 4.330972671508789,
+      "eval_runtime": 33.6382,
+      "eval_samples_per_second": 0.743,
+      "eval_steps_per_second": 0.743,
+      "step": 7488
+    },
+    {
+      "epoch": 32.00008547008547,
+      "grad_norm": 0.0023284293711185455,
+      "learning_rate": 7.554605887939222e-06,
+      "loss": 0.2603,
+      "step": 7490
+    },
+    {
+      "epoch": 32.00051282051282,
+      "grad_norm": 0.01587550714612007,
+      "learning_rate": 7.54985754985755e-06,
+      "loss": 0.1526,
+      "step": 7500
+    },
+    {
+      "epoch": 32.00094017094017,
+      "grad_norm": 345.5118408203125,
+      "learning_rate": 7.545109211775879e-06,
+      "loss": 0.1876,
+      "step": 7510
+    },
+    {
+      "epoch": 32.00136752136752,
+      "grad_norm": 0.13611197471618652,
+      "learning_rate": 7.540360873694208e-06,
+      "loss": 0.0039,
+      "step": 7520
+    },
+    {
+      "epoch": 32.00179487179487,
+      "grad_norm": 378.3825988769531,
+      "learning_rate": 7.535612535612537e-06,
+      "loss": 0.5279,
+      "step": 7530
+    },
+    {
+      "epoch": 32.00222222222222,
+      "grad_norm": 0.00310819735750556,
+      "learning_rate": 7.530864197530865e-06,
+      "loss": 0.0143,
+      "step": 7540
+    },
+    {
+      "epoch": 32.002649572649574,
+      "grad_norm": 0.009138006716966629,
+      "learning_rate": 7.526115859449193e-06,
+      "loss": 0.0012,
+      "step": 7550
+    },
+    {
+      "epoch": 32.003076923076925,
+      "grad_norm": 96.70939636230469,
+      "learning_rate": 7.521367521367522e-06,
+      "loss": 0.8369,
+      "step": 7560
+    },
+    {
+      "epoch": 32.00350427350428,
+      "grad_norm": 136.33778381347656,
+      "learning_rate": 7.516619183285851e-06,
+      "loss": 0.5351,
+      "step": 7570
+    },
+    {
+      "epoch": 32.00393162393162,
+      "grad_norm": 0.0020322606433182955,
+      "learning_rate": 7.511870845204179e-06,
+      "loss": 1.2159,
+      "step": 7580
+    },
+    {
+      "epoch": 32.00435897435897,
+      "grad_norm": 0.03725990280508995,
+      "learning_rate": 7.507122507122507e-06,
+      "loss": 0.6448,
+      "step": 7590
+    },
+    {
+      "epoch": 32.00478632478632,
+      "grad_norm": 0.0038324042689055204,
+      "learning_rate": 7.502374169040836e-06,
+      "loss": 0.5215,
+      "step": 7600
+    },
+    {
+      "epoch": 32.005213675213675,
+      "grad_norm": 0.009491757489740849,
+      "learning_rate": 7.497625830959166e-06,
+      "loss": 0.8682,
+      "step": 7610
+    },
+    {
+      "epoch": 32.005641025641026,
+      "grad_norm": 0.005307774059474468,
+      "learning_rate": 7.492877492877494e-06,
+      "loss": 0.0158,
+      "step": 7620
+    },
+    {
+      "epoch": 32.00606837606838,
+      "grad_norm": 0.004275487270206213,
+      "learning_rate": 7.488129154795822e-06,
+      "loss": 0.5831,
+      "step": 7630
+    },
+    {
+      "epoch": 32.00649572649573,
+      "grad_norm": 0.011664590798318386,
+      "learning_rate": 7.4833808167141505e-06,
+      "loss": 0.0162,
+      "step": 7640
+    },
+    {
+      "epoch": 32.00692307692308,
+      "grad_norm": 554.8225708007812,
+      "learning_rate": 7.47863247863248e-06,
+      "loss": 0.2277,
+      "step": 7650
+    },
+    {
+      "epoch": 32.007350427350424,
+      "grad_norm": 0.0021805327851325274,
+      "learning_rate": 7.473884140550808e-06,
+      "loss": 2.079,
+      "step": 7660
+    },
+    {
+      "epoch": 32.007777777777775,
+      "grad_norm": 2.559633255004883,
+      "learning_rate": 7.469135802469136e-06,
+      "loss": 0.0108,
+      "step": 7670
+    },
+    {
+      "epoch": 32.00820512820513,
+      "grad_norm": 15.645346641540527,
+      "learning_rate": 7.4643874643874645e-06,
+      "loss": 0.9423,
+      "step": 7680
+    },
+    {
+      "epoch": 32.00863247863248,
+      "grad_norm": 0.0020319002214819193,
+      "learning_rate": 7.459639126305793e-06,
+      "loss": 0.7076,
+      "step": 7690
+    },
+    {
+      "epoch": 32.00905982905983,
+      "grad_norm": 0.6667153239250183,
+      "learning_rate": 7.454890788224122e-06,
+      "loss": 0.5031,
+      "step": 7700
+    },
+    {
+      "epoch": 32.00948717948718,
+      "grad_norm": 0.0021193181164562702,
+      "learning_rate": 7.450142450142451e-06,
+      "loss": 0.3348,
+      "step": 7710
+    },
+    {
+      "epoch": 32.00991452991453,
+      "grad_norm": 0.005187495611608028,
+      "learning_rate": 7.445394112060779e-06,
+      "loss": 0.0005,
+      "step": 7720
+    },
+    {
+      "epoch": 32.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 4.535378932952881,
+      "eval_runtime": 34.0349,
+      "eval_samples_per_second": 0.735,
+      "eval_steps_per_second": 0.735,
+      "step": 7722
+    },
+    {
+      "epoch": 33.00034188034188,
+      "grad_norm": 0.0027542465832084417,
+      "learning_rate": 7.4406457739791086e-06,
+      "loss": 0.267,
+      "step": 7730
+    },
+    {
+      "epoch": 33.00076923076923,
+      "grad_norm": 0.00943476427346468,
+      "learning_rate": 7.435897435897437e-06,
+      "loss": 0.4691,
+      "step": 7740
+    },
+    {
+      "epoch": 33.00119658119658,
+      "grad_norm": 96.7542724609375,
+      "learning_rate": 7.431149097815765e-06,
+      "loss": 0.1307,
+      "step": 7750
+    },
+    {
+      "epoch": 33.00162393162393,
+      "grad_norm": 0.005492000840604305,
+      "learning_rate": 7.4264007597340934e-06,
+      "loss": 0.0305,
+      "step": 7760
+    },
+    {
+      "epoch": 33.00205128205128,
+      "grad_norm": 12.753386497497559,
+      "learning_rate": 7.421652421652423e-06,
+      "loss": 0.5948,
+      "step": 7770
+    },
+    {
+      "epoch": 33.002478632478635,
+      "grad_norm": 0.0043577225878834724,
+      "learning_rate": 7.416904083570751e-06,
+      "loss": 0.6122,
+      "step": 7780
+    },
+    {
+      "epoch": 33.002905982905986,
+      "grad_norm": 0.00950498878955841,
+      "learning_rate": 7.412155745489079e-06,
+      "loss": 0.1552,
+      "step": 7790
+    },
+    {
+      "epoch": 33.00333333333333,
+      "grad_norm": 146.59999084472656,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.7224,
+      "step": 7800
+    },
+    {
+      "epoch": 33.00376068376068,
+      "grad_norm": 0.0066252099350094795,
+      "learning_rate": 7.402659069325736e-06,
+      "loss": 0.0714,
+      "step": 7810
+    },
+    {
+      "epoch": 33.00418803418803,
+      "grad_norm": 0.003962312359362841,
+      "learning_rate": 7.397910731244066e-06,
+      "loss": 0.1235,
+      "step": 7820
+    },
+    {
+      "epoch": 33.004615384615384,
+      "grad_norm": 0.004090788774192333,
+      "learning_rate": 7.393162393162394e-06,
+      "loss": 0.8119,
+      "step": 7830
+    },
+    {
+      "epoch": 33.005042735042736,
+      "grad_norm": 0.005719439126551151,
+      "learning_rate": 7.388414055080722e-06,
+      "loss": 1.4667,
+      "step": 7840
+    },
+    {
+      "epoch": 33.00547008547009,
+      "grad_norm": 0.004016694147139788,
+      "learning_rate": 7.3836657169990515e-06,
+      "loss": 1.7774,
+      "step": 7850
+    },
+    {
+      "epoch": 33.00589743589744,
+      "grad_norm": 0.02771218866109848,
+      "learning_rate": 7.37891737891738e-06,
+      "loss": 0.1237,
+      "step": 7860
+    },
+    {
+      "epoch": 33.00632478632479,
+      "grad_norm": 10.215202331542969,
+      "learning_rate": 7.374169040835708e-06,
+      "loss": 0.2107,
+      "step": 7870
+    },
+    {
+      "epoch": 33.006752136752134,
+      "grad_norm": 0.036797523498535156,
+      "learning_rate": 7.369420702754036e-06,
+      "loss": 0.3696,
+      "step": 7880
+    },
+    {
+      "epoch": 33.007179487179485,
+      "grad_norm": 0.0023336284793913364,
+      "learning_rate": 7.364672364672365e-06,
+      "loss": 0.0944,
+      "step": 7890
+    },
+    {
+      "epoch": 33.007606837606836,
+      "grad_norm": 0.002150049665942788,
+      "learning_rate": 7.359924026590694e-06,
+      "loss": 0.1512,
+      "step": 7900
+    },
+    {
+      "epoch": 33.00803418803419,
+      "grad_norm": 0.027237646281719208,
+      "learning_rate": 7.355175688509022e-06,
+      "loss": 0.1805,
+      "step": 7910
+    },
+    {
+      "epoch": 33.00846153846154,
+      "grad_norm": 0.004415407776832581,
+      "learning_rate": 7.350427350427351e-06,
+      "loss": 0.6579,
+      "step": 7920
+    },
+    {
+      "epoch": 33.00888888888889,
+      "grad_norm": 0.009564803913235664,
+      "learning_rate": 7.34567901234568e-06,
+      "loss": 0.6692,
+      "step": 7930
+    },
+    {
+      "epoch": 33.00931623931624,
+      "grad_norm": 0.012014131061732769,
+      "learning_rate": 7.340930674264009e-06,
+      "loss": 0.0433,
+      "step": 7940
+    },
+    {
+      "epoch": 33.00974358974359,
+      "grad_norm": 1.3320039510726929,
+      "learning_rate": 7.336182336182337e-06,
+      "loss": 0.004,
+      "step": 7950
+    },
+    {
+      "epoch": 33.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 4.597040176391602,
+      "eval_runtime": 33.0791,
+      "eval_samples_per_second": 0.756,
+      "eval_steps_per_second": 0.756,
+      "step": 7956
+    },
+    {
+      "epoch": 34.00017094017094,
+      "grad_norm": 0.011348673142492771,
+      "learning_rate": 7.331433998100665e-06,
+      "loss": 0.0008,
+      "step": 7960
+    },
+    {
+      "epoch": 34.00059829059829,
+      "grad_norm": 0.008984715677797794,
+      "learning_rate": 7.326685660018994e-06,
+      "loss": 0.026,
+      "step": 7970
+    },
+    {
+      "epoch": 34.00102564102564,
+      "grad_norm": 0.002917045494541526,
+      "learning_rate": 7.321937321937323e-06,
+      "loss": 1.1443,
+      "step": 7980
+    },
+    {
+      "epoch": 34.00145299145299,
+      "grad_norm": 215.349365234375,
+      "learning_rate": 7.317188983855651e-06,
+      "loss": 0.4194,
+      "step": 7990
+    },
+    {
+      "epoch": 34.001880341880344,
+      "grad_norm": 0.02057277224957943,
+      "learning_rate": 7.312440645773979e-06,
+      "loss": 0.3795,
+      "step": 8000
+    },
+    {
+      "epoch": 34.002307692307696,
+      "grad_norm": 0.009319877251982689,
+      "learning_rate": 7.307692307692308e-06,
+      "loss": 1.1649,
+      "step": 8010
+    },
+    {
+      "epoch": 34.00273504273504,
+      "grad_norm": 0.003309460124000907,
+      "learning_rate": 7.302943969610636e-06,
+      "loss": 0.665,
+      "step": 8020
+    },
+    {
+      "epoch": 34.00316239316239,
+      "grad_norm": 0.0023234267719089985,
+      "learning_rate": 7.298195631528966e-06,
+      "loss": 0.1577,
+      "step": 8030
+    },
+    {
+      "epoch": 34.00358974358974,
+      "grad_norm": 19.653993606567383,
+      "learning_rate": 7.293447293447294e-06,
+      "loss": 1.0529,
+      "step": 8040
+    },
+    {
+      "epoch": 34.004017094017094,
+      "grad_norm": 0.0021447527687996626,
+      "learning_rate": 7.2886989553656225e-06,
+      "loss": 0.6675,
+      "step": 8050
+    },
+    {
+      "epoch": 34.004444444444445,
+      "grad_norm": 0.011748022399842739,
+      "learning_rate": 7.283950617283952e-06,
+      "loss": 0.0002,
+      "step": 8060
+    },
+    {
+      "epoch": 34.0048717948718,
+      "grad_norm": 0.0019833322148770094,
+      "learning_rate": 7.27920227920228e-06,
+      "loss": 0.5008,
+      "step": 8070
+    },
+    {
+      "epoch": 34.00529914529915,
+      "grad_norm": 0.016223294660449028,
+      "learning_rate": 7.274453941120608e-06,
+      "loss": 0.0013,
+      "step": 8080
+    },
+    {
+      "epoch": 34.0057264957265,
+      "grad_norm": 0.001770445262081921,
+      "learning_rate": 7.2697056030389366e-06,
+      "loss": 0.761,
+      "step": 8090
+    },
+    {
+      "epoch": 34.00615384615384,
+      "grad_norm": 0.005002091638743877,
+      "learning_rate": 7.264957264957266e-06,
+      "loss": 0.0004,
+      "step": 8100
+    },
+    {
+      "epoch": 34.006581196581195,
+      "grad_norm": 0.0015637052711099386,
+      "learning_rate": 7.260208926875594e-06,
+      "loss": 1.2961,
+      "step": 8110
+    },
+    {
+      "epoch": 34.007008547008546,
+      "grad_norm": 0.002154473215341568,
+      "learning_rate": 7.255460588793922e-06,
+      "loss": 0.0005,
+      "step": 8120
+    },
+    {
+      "epoch": 34.0074358974359,
+      "grad_norm": 0.004174029920250177,
+      "learning_rate": 7.2507122507122514e-06,
+      "loss": 0.2425,
+      "step": 8130
+    },
+    {
+      "epoch": 34.00786324786325,
+      "grad_norm": 0.0016593949403613806,
+      "learning_rate": 7.245963912630581e-06,
+      "loss": 0.7072,
+      "step": 8140
+    },
+    {
+      "epoch": 34.0082905982906,
+      "grad_norm": 0.003939433954656124,
+      "learning_rate": 7.241215574548909e-06,
+      "loss": 0.2379,
+      "step": 8150
+    },
+    {
+      "epoch": 34.00871794871795,
+      "grad_norm": 0.0038190174382179976,
+      "learning_rate": 7.236467236467237e-06,
+      "loss": 0.0005,
+      "step": 8160
+    },
+    {
+      "epoch": 34.0091452991453,
+      "grad_norm": 0.009816322475671768,
+      "learning_rate": 7.2317188983855655e-06,
+      "loss": 0.0006,
+      "step": 8170
+    },
+    {
+      "epoch": 34.00957264957265,
+      "grad_norm": 0.0018094313563778996,
+      "learning_rate": 7.226970560303895e-06,
+      "loss": 0.0108,
+      "step": 8180
+    },
+    {
+      "epoch": 34.01,
+      "grad_norm": 0.0064643691293895245,
+      "learning_rate": 7.222222222222223e-06,
+      "loss": 0.3017,
+      "step": 8190
+    },
+    {
+      "epoch": 34.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 4.587911605834961,
+      "eval_runtime": 33.8673,
+      "eval_samples_per_second": 0.738,
+      "eval_steps_per_second": 0.738,
+      "step": 8190
+    },
+    {
+      "epoch": 35.00042735042735,
+      "grad_norm": 0.00291146500967443,
+      "learning_rate": 7.217473884140551e-06,
+      "loss": 0.002,
+      "step": 8200
+    },
+    {
+      "epoch": 35.0008547008547,
+      "grad_norm": 0.009560974314808846,
+      "learning_rate": 7.2127255460588795e-06,
+      "loss": 0.0003,
+      "step": 8210
+    },
+    {
+      "epoch": 35.001282051282054,
+      "grad_norm": 66.06437683105469,
+      "learning_rate": 7.207977207977208e-06,
+      "loss": 1.0965,
+      "step": 8220
+    },
+    {
+      "epoch": 35.001709401709405,
+      "grad_norm": 0.0027835587970912457,
+      "learning_rate": 7.203228869895537e-06,
+      "loss": 0.0116,
+      "step": 8230
+    },
+    {
+      "epoch": 35.00213675213675,
+      "grad_norm": 0.022427700459957123,
+      "learning_rate": 7.198480531813866e-06,
+      "loss": 0.4445,
+      "step": 8240
+    },
+    {
+      "epoch": 35.0025641025641,
+      "grad_norm": 6.788797378540039,
+      "learning_rate": 7.193732193732194e-06,
+      "loss": 0.7864,
+      "step": 8250
+    },
+    {
+      "epoch": 35.00299145299145,
+      "grad_norm": 0.002020070794969797,
+      "learning_rate": 7.1889838556505235e-06,
+      "loss": 0.7858,
+      "step": 8260
+    },
+    {
+      "epoch": 35.0034188034188,
+      "grad_norm": 0.08197605609893799,
+      "learning_rate": 7.184235517568852e-06,
+      "loss": 0.6455,
+      "step": 8270
+    },
+    {
+      "epoch": 35.003846153846155,
+      "grad_norm": 0.001466662622988224,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 0.057,
+      "step": 8280
+    },
+    {
+      "epoch": 35.004273504273506,
+      "grad_norm": 0.014914972707629204,
+      "learning_rate": 7.174738841405508e-06,
+      "loss": 0.065,
+      "step": 8290
+    },
+    {
+      "epoch": 35.00470085470086,
+      "grad_norm": 0.04162408038973808,
+      "learning_rate": 7.169990503323837e-06,
+      "loss": 0.0734,
+      "step": 8300
+    },
+    {
+      "epoch": 35.00512820512821,
+      "grad_norm": 0.33848291635513306,
+      "learning_rate": 7.165242165242166e-06,
+      "loss": 0.7488,
+      "step": 8310
+    },
+    {
+      "epoch": 35.00555555555555,
+      "grad_norm": 0.00879302341490984,
+      "learning_rate": 7.160493827160494e-06,
+      "loss": 0.0036,
+      "step": 8320
+    },
+    {
+      "epoch": 35.005982905982904,
+      "grad_norm": 0.0014278884045779705,
+      "learning_rate": 7.1557454890788224e-06,
+      "loss": 0.8021,
+      "step": 8330
+    },
+    {
+      "epoch": 35.006410256410255,
+      "grad_norm": 0.03050161898136139,
+      "learning_rate": 7.1509971509971524e-06,
+      "loss": 0.0002,
+      "step": 8340
+    },
+    {
+      "epoch": 35.00683760683761,
+      "grad_norm": 0.0032293125987052917,
+      "learning_rate": 7.146248812915481e-06,
+      "loss": 0.4796,
+      "step": 8350
+    },
+    {
+      "epoch": 35.00726495726496,
+      "grad_norm": 0.0025629340671002865,
+      "learning_rate": 7.141500474833809e-06,
+      "loss": 0.7034,
+      "step": 8360
+    },
+    {
+      "epoch": 35.00769230769231,
+      "grad_norm": 0.004474421963095665,
+      "learning_rate": 7.136752136752137e-06,
+      "loss": 0.0047,
+      "step": 8370
+    },
+    {
+      "epoch": 35.00811965811966,
+      "grad_norm": 0.04403112828731537,
+      "learning_rate": 7.132003798670466e-06,
+      "loss": 0.9985,
+      "step": 8380
+    },
+    {
+      "epoch": 35.00854700854701,
+      "grad_norm": 0.002210602629929781,
+      "learning_rate": 7.127255460588795e-06,
+      "loss": 0.0011,
+      "step": 8390
+    },
+    {
+      "epoch": 35.008974358974356,
+      "grad_norm": 413.6798400878906,
+      "learning_rate": 7.122507122507123e-06,
+      "loss": 0.378,
+      "step": 8400
+    },
+    {
+      "epoch": 35.00940170940171,
+      "grad_norm": 0.003414222039282322,
+      "learning_rate": 7.117758784425451e-06,
+      "loss": 0.722,
+      "step": 8410
+    },
+    {
+      "epoch": 35.00982905982906,
+      "grad_norm": 0.0017525185830891132,
+      "learning_rate": 7.11301044634378e-06,
+      "loss": 0.2014,
+      "step": 8420
+    },
+    {
+      "epoch": 35.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 4.280939102172852,
+      "eval_runtime": 33.6967,
+      "eval_samples_per_second": 0.742,
+      "eval_steps_per_second": 0.742,
+      "step": 8424
+    },
+    {
+      "epoch": 36.00025641025641,
+      "grad_norm": 323.9665222167969,
+      "learning_rate": 7.108262108262109e-06,
+      "loss": 0.7504,
+      "step": 8430
+    },
+    {
+      "epoch": 36.00068376068376,
+      "grad_norm": 0.0009274838957935572,
+      "learning_rate": 7.103513770180438e-06,
+      "loss": 0.0036,
+      "step": 8440
+    },
+    {
+      "epoch": 36.00111111111111,
+      "grad_norm": 0.004535932093858719,
+      "learning_rate": 7.098765432098766e-06,
+      "loss": 0.0003,
+      "step": 8450
+    },
+    {
+      "epoch": 36.00153846153846,
+      "grad_norm": 0.007801331579685211,
+      "learning_rate": 7.0940170940170945e-06,
+      "loss": 0.0066,
+      "step": 8460
+    },
+    {
+      "epoch": 36.00196581196581,
+      "grad_norm": 1.9428983926773071,
+      "learning_rate": 7.089268755935424e-06,
+      "loss": 0.5418,
+      "step": 8470
+    },
+    {
+      "epoch": 36.00239316239316,
+      "grad_norm": 0.09623009711503983,
+      "learning_rate": 7.084520417853752e-06,
+      "loss": 0.2488,
+      "step": 8480
+    },
+    {
+      "epoch": 36.00282051282051,
+      "grad_norm": 0.020373547449707985,
+      "learning_rate": 7.07977207977208e-06,
+      "loss": 0.0012,
+      "step": 8490
+    },
+    {
+      "epoch": 36.003247863247864,
+      "grad_norm": 0.002565371571108699,
+      "learning_rate": 7.0750237416904086e-06,
+      "loss": 0.0004,
+      "step": 8500
+    },
+    {
+      "epoch": 36.003675213675216,
+      "grad_norm": 243.3353271484375,
+      "learning_rate": 7.070275403608738e-06,
+      "loss": 0.1137,
+      "step": 8510
+    },
+    {
+      "epoch": 36.00410256410257,
+      "grad_norm": 0.0050058732740581036,
+      "learning_rate": 7.065527065527066e-06,
+      "loss": 0.7545,
+      "step": 8520
+    },
+    {
+      "epoch": 36.00452991452991,
+      "grad_norm": 0.0049826339818537235,
+      "learning_rate": 7.060778727445394e-06,
+      "loss": 0.0002,
+      "step": 8530
+    },
+    {
+      "epoch": 36.00495726495726,
+      "grad_norm": 0.0015328944427892566,
+      "learning_rate": 7.056030389363723e-06,
+      "loss": 1.5905,
+      "step": 8540
+    },
+    {
+      "epoch": 36.005384615384614,
+      "grad_norm": 0.12297794222831726,
+      "learning_rate": 7.051282051282053e-06,
+      "loss": 0.0142,
+      "step": 8550
+    },
+    {
+      "epoch": 36.005811965811965,
+      "grad_norm": 363.6435546875,
+      "learning_rate": 7.046533713200381e-06,
+      "loss": 0.4281,
+      "step": 8560
+    },
+    {
+      "epoch": 36.006239316239316,
+      "grad_norm": 10.307737350463867,
+      "learning_rate": 7.041785375118709e-06,
+      "loss": 0.0332,
+      "step": 8570
+    },
+    {
+      "epoch": 36.00666666666667,
+      "grad_norm": 0.001833245623856783,
+      "learning_rate": 7.0370370370370375e-06,
+      "loss": 0.1285,
+      "step": 8580
+    },
+    {
+      "epoch": 36.00709401709402,
+      "grad_norm": 75.9461441040039,
+      "learning_rate": 7.032288698955367e-06,
+      "loss": 0.0096,
+      "step": 8590
+    },
+    {
+      "epoch": 36.00752136752137,
+      "grad_norm": 0.007748633157461882,
+      "learning_rate": 7.027540360873695e-06,
+      "loss": 0.0074,
+      "step": 8600
+    },
+    {
+      "epoch": 36.007948717948715,
+      "grad_norm": 0.0246294979006052,
+      "learning_rate": 7.022792022792023e-06,
+      "loss": 0.0058,
+      "step": 8610
+    },
+    {
+      "epoch": 36.008376068376066,
+      "grad_norm": 0.07270914316177368,
+      "learning_rate": 7.0180436847103515e-06,
+      "loss": 0.0007,
+      "step": 8620
+    },
+    {
+      "epoch": 36.00880341880342,
+      "grad_norm": 0.0704275518655777,
+      "learning_rate": 7.01329534662868e-06,
+      "loss": 0.0986,
+      "step": 8630
+    },
+    {
+      "epoch": 36.00923076923077,
+      "grad_norm": 0.03107783943414688,
+      "learning_rate": 7.008547008547009e-06,
+      "loss": 0.0033,
+      "step": 8640
+    },
+    {
+      "epoch": 36.00965811965812,
+      "grad_norm": 0.21083621680736542,
+      "learning_rate": 7.003798670465338e-06,
+      "loss": 0.1573,
+      "step": 8650
+    },
+    {
+      "epoch": 36.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 4.6822123527526855,
+      "eval_runtime": 33.5951,
+      "eval_samples_per_second": 0.744,
+      "eval_steps_per_second": 0.744,
+      "step": 8658
+    },
+    {
+      "epoch": 37.00008547008547,
+      "grad_norm": 318.6376953125,
+      "learning_rate": 6.999050332383666e-06,
+      "loss": 0.5961,
+      "step": 8660
+    },
+    {
+      "epoch": 37.00051282051282,
+      "grad_norm": 1.5630220174789429,
+      "learning_rate": 6.9943019943019955e-06,
+      "loss": 0.005,
+      "step": 8670
+    },
+    {
+      "epoch": 37.00094017094017,
+      "grad_norm": 0.00120359449647367,
+      "learning_rate": 6.989553656220324e-06,
+      "loss": 0.0008,
+      "step": 8680
+    },
+    {
+      "epoch": 37.00136752136752,
+      "grad_norm": 0.0014155134558677673,
+      "learning_rate": 6.984805318138652e-06,
+      "loss": 0.4491,
+      "step": 8690
+    },
+    {
+      "epoch": 37.00179487179487,
+      "grad_norm": 0.0017230873927474022,
+      "learning_rate": 6.9800569800569804e-06,
+      "loss": 0.0004,
+      "step": 8700
+    },
+    {
+      "epoch": 37.00222222222222,
+      "grad_norm": 0.07810331135988235,
+      "learning_rate": 6.975308641975309e-06,
+      "loss": 0.0007,
+      "step": 8710
+    },
+    {
+      "epoch": 37.002649572649574,
+      "grad_norm": 0.0017077282536774874,
+      "learning_rate": 6.970560303893638e-06,
+      "loss": 0.0988,
+      "step": 8720
+    },
+    {
+      "epoch": 37.003076923076925,
+      "grad_norm": 0.004637227393686771,
+      "learning_rate": 6.965811965811966e-06,
+      "loss": 0.1457,
+      "step": 8730
+    },
+    {
+      "epoch": 37.00350427350428,
+      "grad_norm": 0.017330659553408623,
+      "learning_rate": 6.9610636277302945e-06,
+      "loss": 0.0003,
+      "step": 8740
+    },
+    {
+      "epoch": 37.00393162393162,
+      "grad_norm": 0.0017195155378431082,
+      "learning_rate": 6.956315289648623e-06,
+      "loss": 1.9324,
+      "step": 8750
+    },
+    {
+      "epoch": 37.00435897435897,
+      "grad_norm": 0.0010282599832862616,
+      "learning_rate": 6.951566951566953e-06,
+      "loss": 0.6652,
+      "step": 8760
+    },
+    {
+      "epoch": 37.00478632478632,
+      "grad_norm": 28.446561813354492,
+      "learning_rate": 6.946818613485281e-06,
+      "loss": 0.9279,
+      "step": 8770
+    },
+    {
+      "epoch": 37.005213675213675,
+      "grad_norm": 0.022886212915182114,
+      "learning_rate": 6.942070275403609e-06,
+      "loss": 0.9287,
+      "step": 8780
+    },
+    {
+      "epoch": 37.005641025641026,
+      "grad_norm": 0.006739874370396137,
+      "learning_rate": 6.937321937321938e-06,
+      "loss": 0.0193,
+      "step": 8790
+    },
+    {
+      "epoch": 37.00606837606838,
+      "grad_norm": 0.0022217826917767525,
+      "learning_rate": 6.932573599240267e-06,
+      "loss": 0.3743,
+      "step": 8800
+    },
+    {
+      "epoch": 37.00649572649573,
+      "grad_norm": 353.1299133300781,
+      "learning_rate": 6.927825261158595e-06,
+      "loss": 0.7619,
+      "step": 8810
+    },
+    {
+      "epoch": 37.00692307692308,
+      "grad_norm": 0.010308923199772835,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 0.5226,
+      "step": 8820
+    },
+    {
+      "epoch": 37.007350427350424,
+      "grad_norm": 0.386707603931427,
+      "learning_rate": 6.918328584995252e-06,
+      "loss": 0.6146,
+      "step": 8830
+    },
+    {
+      "epoch": 37.007777777777775,
+      "grad_norm": 0.01583043485879898,
+      "learning_rate": 6.913580246913581e-06,
+      "loss": 1.2693,
+      "step": 8840
+    },
+    {
+      "epoch": 37.00820512820513,
+      "grad_norm": 0.41986873745918274,
+      "learning_rate": 6.908831908831909e-06,
+      "loss": 0.4624,
+      "step": 8850
+    },
+    {
+      "epoch": 37.00863247863248,
+      "grad_norm": 0.05952145531773567,
+      "learning_rate": 6.904083570750238e-06,
+      "loss": 0.1098,
+      "step": 8860
+    },
+    {
+      "epoch": 37.00905982905983,
+      "grad_norm": 0.5341818332672119,
+      "learning_rate": 6.8993352326685666e-06,
+      "loss": 0.0011,
+      "step": 8870
+    },
+    {
+      "epoch": 37.00948717948718,
+      "grad_norm": 0.7262808680534363,
+      "learning_rate": 6.894586894586896e-06,
+      "loss": 0.1184,
+      "step": 8880
+    },
+    {
+      "epoch": 37.00991452991453,
+      "grad_norm": 0.006815705914050341,
+      "learning_rate": 6.889838556505224e-06,
+      "loss": 0.0041,
+      "step": 8890
+    },
+    {
+      "epoch": 37.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 5.167320728302002,
+      "eval_runtime": 33.711,
+      "eval_samples_per_second": 0.742,
+      "eval_steps_per_second": 0.742,
+      "step": 8892
+    },
+    {
+      "epoch": 38.00034188034188,
+      "grad_norm": 0.0014624128816649318,
+      "learning_rate": 6.885090218423552e-06,
+      "loss": 0.0327,
+      "step": 8900
+    },
+    {
+      "epoch": 38.00076923076923,
+      "grad_norm": 0.4525696933269501,
+      "learning_rate": 6.880341880341881e-06,
+      "loss": 0.0079,
+      "step": 8910
+    },
+    {
+      "epoch": 38.00119658119658,
+      "grad_norm": 0.0014630717923864722,
+      "learning_rate": 6.87559354226021e-06,
+      "loss": 0.0003,
+      "step": 8920
+    },
+    {
+      "epoch": 38.00162393162393,
+      "grad_norm": 0.009966408833861351,
+      "learning_rate": 6.870845204178538e-06,
+      "loss": 0.0023,
+      "step": 8930
+    },
+    {
+      "epoch": 38.00205128205128,
+      "grad_norm": 0.02271059900522232,
+      "learning_rate": 6.866096866096866e-06,
+      "loss": 0.0006,
+      "step": 8940
+    },
+    {
+      "epoch": 38.002478632478635,
+      "grad_norm": 0.15712451934814453,
+      "learning_rate": 6.861348528015195e-06,
+      "loss": 0.0004,
+      "step": 8950
+    },
+    {
+      "epoch": 38.002905982905986,
+      "grad_norm": 0.011041790246963501,
+      "learning_rate": 6.856600189933523e-06,
+      "loss": 1.3034,
+      "step": 8960
+    },
+    {
+      "epoch": 38.00333333333333,
+      "grad_norm": 0.0015798502136021852,
+      "learning_rate": 6.851851851851853e-06,
+      "loss": 0.0009,
+      "step": 8970
+    },
+    {
+      "epoch": 38.00376068376068,
+      "grad_norm": 0.008027788251638412,
+      "learning_rate": 6.847103513770181e-06,
+      "loss": 0.4744,
+      "step": 8980
+    },
+    {
+      "epoch": 38.00418803418803,
+      "grad_norm": 0.045825887471437454,
+      "learning_rate": 6.8423551756885095e-06,
+      "loss": 0.003,
+      "step": 8990
+    },
+    {
+      "epoch": 38.004615384615384,
+      "grad_norm": 0.0023124797735363245,
+      "learning_rate": 6.837606837606839e-06,
+      "loss": 0.0006,
+      "step": 9000
+    },
+    {
+      "epoch": 38.005042735042736,
+      "grad_norm": 4.002900123596191,
+      "learning_rate": 6.832858499525167e-06,
+      "loss": 0.5629,
+      "step": 9010
+    },
+    {
+      "epoch": 38.00547008547009,
+      "grad_norm": 0.1215251013636589,
+      "learning_rate": 6.828110161443495e-06,
+      "loss": 0.0034,
+      "step": 9020
+    },
+    {
+      "epoch": 38.00589743589744,
+      "grad_norm": 0.007816506549715996,
+      "learning_rate": 6.8233618233618235e-06,
+      "loss": 0.3739,
+      "step": 9030
+    },
+    {
+      "epoch": 38.00632478632479,
+      "grad_norm": 80.12126922607422,
+      "learning_rate": 6.818613485280152e-06,
+      "loss": 0.258,
+      "step": 9040
+    },
+    {
+      "epoch": 38.006752136752134,
+      "grad_norm": 0.0014129136689007282,
+      "learning_rate": 6.813865147198481e-06,
+      "loss": 0.0002,
+      "step": 9050
+    },
+    {
+      "epoch": 38.007179487179485,
+      "grad_norm": 507.7630310058594,
+      "learning_rate": 6.809116809116809e-06,
+      "loss": 0.2478,
+      "step": 9060
+    },
+    {
+      "epoch": 38.007606837606836,
+      "grad_norm": 0.0028367233462631702,
+      "learning_rate": 6.804368471035138e-06,
+      "loss": 0.5826,
+      "step": 9070
+    },
+    {
+      "epoch": 38.00803418803419,
+      "grad_norm": 0.018132785335183144,
+      "learning_rate": 6.7996201329534676e-06,
+      "loss": 0.3636,
+      "step": 9080
+    },
+    {
+      "epoch": 38.00846153846154,
+      "grad_norm": 1.1534035205841064,
+      "learning_rate": 6.794871794871796e-06,
+      "loss": 0.454,
+      "step": 9090
+    },
+    {
+      "epoch": 38.00888888888889,
+      "grad_norm": 0.0010121689410880208,
+      "learning_rate": 6.790123456790124e-06,
+      "loss": 0.8389,
+      "step": 9100
+    },
+    {
+      "epoch": 38.00931623931624,
+      "grad_norm": 0.0008693256531842053,
+      "learning_rate": 6.7853751187084525e-06,
+      "loss": 0.0002,
+      "step": 9110
+    },
+    {
+      "epoch": 38.00974358974359,
+      "grad_norm": 0.001992259407415986,
+      "learning_rate": 6.780626780626781e-06,
+      "loss": 0.0001,
+      "step": 9120
+    },
+    {
+      "epoch": 38.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 5.400512218475342,
+      "eval_runtime": 33.5959,
+      "eval_samples_per_second": 0.744,
+      "eval_steps_per_second": 0.744,
+      "step": 9126
+    },
+    {
+      "epoch": 39.00017094017094,
+      "grad_norm": 0.010901215486228466,
+      "learning_rate": 6.77587844254511e-06,
+      "loss": 0.0009,
+      "step": 9130
+    },
+    {
+      "epoch": 39.00059829059829,
+      "grad_norm": 0.016072046011686325,
+      "learning_rate": 6.771130104463438e-06,
+      "loss": 0.1631,
+      "step": 9140
+    },
+    {
+      "epoch": 39.00102564102564,
+      "grad_norm": 0.017606910318136215,
+      "learning_rate": 6.7663817663817665e-06,
+      "loss": 0.3611,
+      "step": 9150
+    },
+    {
+      "epoch": 39.00145299145299,
+      "grad_norm": 0.07369554787874222,
+      "learning_rate": 6.761633428300095e-06,
+      "loss": 0.0008,
+      "step": 9160
+    },
+    {
+      "epoch": 39.001880341880344,
+      "grad_norm": 0.023543603718280792,
+      "learning_rate": 6.756885090218425e-06,
+      "loss": 0.121,
+      "step": 9170
+    },
+    {
+      "epoch": 39.002307692307696,
+      "grad_norm": 37.938270568847656,
+      "learning_rate": 6.752136752136753e-06,
+      "loss": 1.2801,
+      "step": 9180
+    },
+    {
+      "epoch": 39.00273504273504,
+      "grad_norm": 0.0017556172097101808,
+      "learning_rate": 6.747388414055081e-06,
+      "loss": 0.0766,
+      "step": 9190
+    },
+    {
+      "epoch": 39.00316239316239,
+      "grad_norm": 0.017493102699518204,
+      "learning_rate": 6.74264007597341e-06,
+      "loss": 0.0422,
+      "step": 9200
+    },
+    {
+      "epoch": 39.00358974358974,
+      "grad_norm": 0.0010195322101935744,
+      "learning_rate": 6.737891737891739e-06,
+      "loss": 0.0576,
+      "step": 9210
+    },
+    {
+      "epoch": 39.004017094017094,
+      "grad_norm": 0.009548988193273544,
+      "learning_rate": 6.733143399810067e-06,
+      "loss": 0.0003,
+      "step": 9220
+    },
+    {
+      "epoch": 39.004444444444445,
+      "grad_norm": 0.00347512518055737,
+      "learning_rate": 6.728395061728395e-06,
+      "loss": 0.0004,
+      "step": 9230
+    },
+    {
+      "epoch": 39.0048717948718,
+      "grad_norm": 0.027547018602490425,
+      "learning_rate": 6.723646723646724e-06,
+      "loss": 0.0002,
+      "step": 9240
+    },
+    {
+      "epoch": 39.00529914529915,
+      "grad_norm": 0.005075276829302311,
+      "learning_rate": 6.718898385565053e-06,
+      "loss": 0.2796,
+      "step": 9250
+    },
+    {
+      "epoch": 39.0057264957265,
+      "grad_norm": 0.03868807852268219,
+      "learning_rate": 6.714150047483381e-06,
+      "loss": 0.005,
+      "step": 9260
+    },
+    {
+      "epoch": 39.00615384615384,
+      "grad_norm": 324.1128234863281,
+      "learning_rate": 6.7094017094017094e-06,
+      "loss": 0.82,
+      "step": 9270
+    },
+    {
+      "epoch": 39.006581196581195,
+      "grad_norm": 0.10437753051519394,
+      "learning_rate": 6.7046533713200394e-06,
+      "loss": 0.9481,
+      "step": 9280
+    },
+    {
+      "epoch": 39.007008547008546,
+      "grad_norm": 0.12134591490030289,
+      "learning_rate": 6.699905033238368e-06,
+      "loss": 0.0065,
+      "step": 9290
+    },
+    {
+      "epoch": 39.0074358974359,
+      "grad_norm": 0.029539357870817184,
+      "learning_rate": 6.695156695156696e-06,
+      "loss": 0.0004,
+      "step": 9300
+    },
+    {
+      "epoch": 39.00786324786325,
+      "grad_norm": 205.1163787841797,
+      "learning_rate": 6.690408357075024e-06,
+      "loss": 0.4614,
+      "step": 9310
+    },
+    {
+      "epoch": 39.0082905982906,
+      "grad_norm": 0.008227716200053692,
+      "learning_rate": 6.685660018993353e-06,
+      "loss": 0.0715,
+      "step": 9320
+    },
+    {
+      "epoch": 39.00871794871795,
+      "grad_norm": 292.14813232421875,
+      "learning_rate": 6.680911680911682e-06,
+      "loss": 0.2026,
+      "step": 9330
+    },
+    {
+      "epoch": 39.0091452991453,
+      "grad_norm": 0.0007633490022271872,
+      "learning_rate": 6.67616334283001e-06,
+      "loss": 0.1259,
+      "step": 9340
+    },
+    {
+      "epoch": 39.00957264957265,
+      "grad_norm": 0.0007212384953163564,
+      "learning_rate": 6.671415004748338e-06,
+      "loss": 0.05,
+      "step": 9350
+    },
+    {
+      "epoch": 39.01,
+      "grad_norm": 0.0017658963333815336,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.1066,
+      "step": 9360
+    },
+    {
+      "epoch": 39.01,
+      "eval_accuracy": 0.48,
+      "eval_loss": 4.4508891105651855,
+      "eval_runtime": 33.5596,
+      "eval_samples_per_second": 0.745,
+      "eval_steps_per_second": 0.745,
+      "step": 9360
+    },
+    {
+      "epoch": 40.00042735042735,
+      "grad_norm": 0.24033397436141968,
+      "learning_rate": 6.661918328584995e-06,
+      "loss": 0.3139,
+      "step": 9370
+    },
+    {
+      "epoch": 40.0008547008547,
+      "grad_norm": 0.11893019825220108,
+      "learning_rate": 6.657169990503325e-06,
+      "loss": 0.07,
+      "step": 9380
+    },
+    {
+      "epoch": 40.001282051282054,
+      "grad_norm": 262.5067443847656,
+      "learning_rate": 6.652421652421653e-06,
+      "loss": 0.4037,
+      "step": 9390
+    },
+    {
+      "epoch": 40.001709401709405,
+      "grad_norm": 0.05252963304519653,
+      "learning_rate": 6.6476733143399815e-06,
+      "loss": 0.3586,
+      "step": 9400
+    },
+    {
+      "epoch": 40.00213675213675,
+      "grad_norm": 255.7079620361328,
+      "learning_rate": 6.642924976258311e-06,
+      "loss": 0.0414,
+      "step": 9410
+    },
+    {
+      "epoch": 40.0025641025641,
+      "grad_norm": 0.07694090157747269,
+      "learning_rate": 6.638176638176639e-06,
+      "loss": 0.3065,
+      "step": 9420
+    },
+    {
+      "epoch": 40.00299145299145,
+      "grad_norm": 0.0017998847179114819,
+      "learning_rate": 6.633428300094967e-06,
+      "loss": 0.0004,
+      "step": 9430
+    },
+    {
+      "epoch": 40.0034188034188,
+      "grad_norm": 208.48390197753906,
+      "learning_rate": 6.6286799620132956e-06,
+      "loss": 0.775,
+      "step": 9440
+    },
+    {
+      "epoch": 40.003846153846155,
+      "grad_norm": 0.004139886237680912,
+      "learning_rate": 6.623931623931624e-06,
+      "loss": 0.0009,
+      "step": 9450
+    },
+    {
+      "epoch": 40.004273504273506,
+      "grad_norm": 0.0015383479185402393,
+      "learning_rate": 6.619183285849953e-06,
+      "loss": 0.0003,
+      "step": 9460
+    },
+    {
+      "epoch": 40.00470085470086,
+      "grad_norm": 79.24043273925781,
+      "learning_rate": 6.614434947768281e-06,
+      "loss": 0.0126,
+      "step": 9470
+    },
+    {
+      "epoch": 40.00512820512821,
+      "grad_norm": 0.10130419582128525,
+      "learning_rate": 6.60968660968661e-06,
+      "loss": 0.0092,
+      "step": 9480
+    },
+    {
+      "epoch": 40.00555555555555,
+      "grad_norm": 0.0018481501610949636,
+      "learning_rate": 6.60493827160494e-06,
+      "loss": 0.0001,
+      "step": 9490
+    },
+    {
+      "epoch": 40.005982905982904,
+      "grad_norm": 0.00359438662417233,
+      "learning_rate": 6.600189933523268e-06,
+      "loss": 0.0015,
+      "step": 9500
+    },
+    {
+      "epoch": 40.006410256410255,
+      "grad_norm": 0.0007249795598909259,
+      "learning_rate": 6.595441595441596e-06,
+      "loss": 0.0001,
+      "step": 9510
+    },
+    {
+      "epoch": 40.00683760683761,
+      "grad_norm": 0.0008260689792223275,
+      "learning_rate": 6.5906932573599245e-06,
+      "loss": 0.5924,
+      "step": 9520
+    },
+    {
+      "epoch": 40.00726495726496,
+      "grad_norm": 0.0014530919725075364,
+      "learning_rate": 6.585944919278253e-06,
+      "loss": 0.7378,
+      "step": 9530
+    },
+    {
+      "epoch": 40.00769230769231,
+      "grad_norm": 14.047000885009766,
+      "learning_rate": 6.581196581196582e-06,
+      "loss": 0.617,
+      "step": 9540
+    },
+    {
+      "epoch": 40.00811965811966,
+      "grad_norm": 0.00833844393491745,
+      "learning_rate": 6.57644824311491e-06,
+      "loss": 0.0337,
+      "step": 9550
+    },
+    {
+      "epoch": 40.00854700854701,
+      "grad_norm": 0.001020289957523346,
+      "learning_rate": 6.5716999050332385e-06,
+      "loss": 0.5866,
+      "step": 9560
+    },
+    {
+      "epoch": 40.008974358974356,
+      "grad_norm": 0.0010084334062412381,
+      "learning_rate": 6.566951566951567e-06,
+      "loss": 0.765,
+      "step": 9570
+    },
+    {
+      "epoch": 40.00940170940171,
+      "grad_norm": 0.0008349449490197003,
+      "learning_rate": 6.562203228869896e-06,
+      "loss": 0.0001,
+      "step": 9580
+    },
+    {
+      "epoch": 40.00982905982906,
+      "grad_norm": 0.003181006060913205,
+      "learning_rate": 6.557454890788225e-06,
+      "loss": 0.0001,
+      "step": 9590
+    },
+    {
+      "epoch": 40.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 5.09063196182251,
+      "eval_runtime": 33.6154,
+      "eval_samples_per_second": 0.744,
+      "eval_steps_per_second": 0.744,
+      "step": 9594
+    },
+    {
+      "epoch": 41.00025641025641,
+      "grad_norm": 0.027484165504574776,
+      "learning_rate": 6.552706552706553e-06,
+      "loss": 0.0001,
+      "step": 9600
+    },
+    {
+      "epoch": 41.00068376068376,
+      "grad_norm": 0.002781185321509838,
+      "learning_rate": 6.5479582146248825e-06,
+      "loss": 0.9085,
+      "step": 9610
+    },
+    {
+      "epoch": 41.00111111111111,
+      "grad_norm": 0.002277638763189316,
+      "learning_rate": 6.543209876543211e-06,
+      "loss": 0.0046,
+      "step": 9620
+    },
+    {
+      "epoch": 41.00153846153846,
+      "grad_norm": 0.001006715465337038,
+      "learning_rate": 6.538461538461539e-06,
+      "loss": 0.2547,
+      "step": 9630
+    },
+    {
+      "epoch": 41.00196581196581,
+      "grad_norm": 0.017884666100144386,
+      "learning_rate": 6.533713200379867e-06,
+      "loss": 0.0003,
+      "step": 9640
+    },
+    {
+      "epoch": 41.00239316239316,
+      "grad_norm": 0.0043096174485981464,
+      "learning_rate": 6.528964862298196e-06,
+      "loss": 0.3413,
+      "step": 9650
+    },
+    {
+      "epoch": 41.00282051282051,
+      "grad_norm": 0.008827520534396172,
+      "learning_rate": 6.524216524216525e-06,
+      "loss": 0.5963,
+      "step": 9660
+    },
+    {
+      "epoch": 41.003247863247864,
+      "grad_norm": 380.2333679199219,
+      "learning_rate": 6.519468186134853e-06,
+      "loss": 0.1933,
+      "step": 9670
+    },
+    {
+      "epoch": 41.003675213675216,
+      "grad_norm": 53.198787689208984,
+      "learning_rate": 6.5147198480531815e-06,
+      "loss": 0.3678,
+      "step": 9680
+    },
+    {
+      "epoch": 41.00410256410257,
+      "grad_norm": 0.0006724594277329743,
+      "learning_rate": 6.50997150997151e-06,
+      "loss": 0.0017,
+      "step": 9690
+    },
+    {
+      "epoch": 41.00452991452991,
+      "grad_norm": 0.0016128338174894452,
+      "learning_rate": 6.50522317188984e-06,
+      "loss": 0.7708,
+      "step": 9700
+    },
+    {
+      "epoch": 41.00495726495726,
+      "grad_norm": 0.0053380681201815605,
+      "learning_rate": 6.500474833808168e-06,
+      "loss": 0.0048,
+      "step": 9710
+    },
+    {
+      "epoch": 41.005384615384614,
+      "grad_norm": 11.402435302734375,
+      "learning_rate": 6.495726495726496e-06,
+      "loss": 0.1091,
+      "step": 9720
+    },
+    {
+      "epoch": 41.005811965811965,
+      "grad_norm": 0.0011073511559516191,
+      "learning_rate": 6.490978157644825e-06,
+      "loss": 0.0594,
+      "step": 9730
+    },
+    {
+      "epoch": 41.006239316239316,
+      "grad_norm": 0.0016850410029292107,
+      "learning_rate": 6.486229819563154e-06,
+      "loss": 0.5871,
+      "step": 9740
+    },
+    {
+      "epoch": 41.00666666666667,
+      "grad_norm": 0.011738145723938942,
+      "learning_rate": 6.481481481481482e-06,
+      "loss": 0.0002,
+      "step": 9750
+    },
+    {
+      "epoch": 41.00709401709402,
+      "grad_norm": 0.02577214501798153,
+      "learning_rate": 6.47673314339981e-06,
+      "loss": 0.7295,
+      "step": 9760
+    },
+    {
+      "epoch": 41.00752136752137,
+      "grad_norm": 0.0008236413705162704,
+      "learning_rate": 6.471984805318139e-06,
+      "loss": 0.0001,
+      "step": 9770
+    },
+    {
+      "epoch": 41.007948717948715,
+      "grad_norm": 0.0007155478815548122,
+      "learning_rate": 6.467236467236467e-06,
+      "loss": 0.2632,
+      "step": 9780
+    },
+    {
+      "epoch": 41.008376068376066,
+      "grad_norm": 0.0013219125103205442,
+      "learning_rate": 6.462488129154796e-06,
+      "loss": 0.0001,
+      "step": 9790
+    },
+    {
+      "epoch": 41.00880341880342,
+      "grad_norm": 203.72512817382812,
+      "learning_rate": 6.457739791073125e-06,
+      "loss": 0.8773,
+      "step": 9800
+    },
+    {
+      "epoch": 41.00923076923077,
+      "grad_norm": 0.0009469674550928175,
+      "learning_rate": 6.4529914529914535e-06,
+      "loss": 0.0001,
+      "step": 9810
+    },
+    {
+      "epoch": 41.00965811965812,
+      "grad_norm": 0.013066309504210949,
+      "learning_rate": 6.448243114909783e-06,
+      "loss": 1.3235,
+      "step": 9820
+    },
+    {
+      "epoch": 41.01,
+      "eval_accuracy": 0.48,
+      "eval_loss": 4.409305572509766,
+      "eval_runtime": 33.7085,
+      "eval_samples_per_second": 0.742,
+      "eval_steps_per_second": 0.742,
+      "step": 9828
+    },
+    {
+      "epoch": 42.00008547008547,
+      "grad_norm": 0.0041074506007134914,
+      "learning_rate": 6.443494776828111e-06,
+      "loss": 0.0003,
+      "step": 9830
+    },
+    {
+      "epoch": 42.00051282051282,
+      "grad_norm": 0.0007542824023403227,
+      "learning_rate": 6.438746438746439e-06,
+      "loss": 0.1122,
+      "step": 9840
+    },
+    {
+      "epoch": 42.00094017094017,
+      "grad_norm": 0.0015057259006425738,
+      "learning_rate": 6.433998100664768e-06,
+      "loss": 0.0002,
+      "step": 9850
+    },
+    {
+      "epoch": 42.00136752136752,
+      "grad_norm": 0.003451643278822303,
+      "learning_rate": 6.429249762583096e-06,
+      "loss": 0.0002,
+      "step": 9860
+    },
+    {
+      "epoch": 42.00179487179487,
+      "grad_norm": 0.0011182624148204923,
+      "learning_rate": 6.424501424501425e-06,
+      "loss": 0.0045,
+      "step": 9870
+    },
+    {
+      "epoch": 42.00222222222222,
+      "grad_norm": 0.0010131365852430463,
+      "learning_rate": 6.419753086419753e-06,
+      "loss": 0.0001,
+      "step": 9880
+    },
+    {
+      "epoch": 42.002649572649574,
+      "grad_norm": 0.0007555413176305592,
+      "learning_rate": 6.415004748338082e-06,
+      "loss": 0.0006,
+      "step": 9890
+    },
+    {
+      "epoch": 42.003076923076925,
+      "grad_norm": 0.0009690074366517365,
+      "learning_rate": 6.410256410256412e-06,
+      "loss": 0.0002,
+      "step": 9900
+    },
+    {
+      "epoch": 42.00350427350428,
+      "grad_norm": 0.0007182178669609129,
+      "learning_rate": 6.40550807217474e-06,
+      "loss": 0.1885,
+      "step": 9910
+    },
+    {
+      "epoch": 42.00393162393162,
+      "grad_norm": 0.0008008884033188224,
+      "learning_rate": 6.400759734093068e-06,
+      "loss": 0.0,
+      "step": 9920
+    },
+    {
+      "epoch": 42.00435897435897,
+      "grad_norm": 0.00729968398809433,
+      "learning_rate": 6.3960113960113965e-06,
+      "loss": 0.9073,
+      "step": 9930
+    },
+    {
+      "epoch": 42.00478632478632,
+      "grad_norm": 0.0006498902221210301,
+      "learning_rate": 6.391263057929726e-06,
+      "loss": 0.0001,
+      "step": 9940
+    },
+    {
+      "epoch": 42.005213675213675,
+      "grad_norm": 0.00066241534659639,
+      "learning_rate": 6.386514719848054e-06,
+      "loss": 1.0737,
+      "step": 9950
+    },
+    {
+      "epoch": 42.005641025641026,
+      "grad_norm": 330.1017150878906,
+      "learning_rate": 6.381766381766382e-06,
+      "loss": 1.9267,
+      "step": 9960
+    },
+    {
+      "epoch": 42.00606837606838,
+      "grad_norm": 0.01651351898908615,
+      "learning_rate": 6.3770180436847105e-06,
+      "loss": 0.0003,
+      "step": 9970
+    },
+    {
+      "epoch": 42.00649572649573,
+      "grad_norm": 0.005164479836821556,
+      "learning_rate": 6.372269705603039e-06,
+      "loss": 0.0008,
+      "step": 9980
+    },
+    {
+      "epoch": 42.00692307692308,
+      "grad_norm": 425.1309814453125,
+      "learning_rate": 6.367521367521368e-06,
+      "loss": 0.3781,
+      "step": 9990
+    },
+    {
+      "epoch": 42.007350427350424,
+      "grad_norm": 0.014773045666515827,
+      "learning_rate": 6.362773029439696e-06,
+      "loss": 0.0002,
+      "step": 10000
+    },
+    {
+      "epoch": 42.007777777777775,
+      "grad_norm": 0.008133858442306519,
+      "learning_rate": 6.358024691358025e-06,
+      "loss": 0.0166,
+      "step": 10010
+    },
+    {
+      "epoch": 42.00820512820513,
+      "grad_norm": 0.4493364691734314,
+      "learning_rate": 6.3532763532763546e-06,
+      "loss": 1.622,
+      "step": 10020
+    },
+    {
+      "epoch": 42.00863247863248,
+      "grad_norm": 0.007051916792988777,
+      "learning_rate": 6.348528015194683e-06,
+      "loss": 0.0094,
+      "step": 10030
+    },
+    {
+      "epoch": 42.00905982905983,
+      "grad_norm": 0.003875425783917308,
+      "learning_rate": 6.343779677113011e-06,
+      "loss": 0.0892,
+      "step": 10040
+    },
+    {
+      "epoch": 42.00948717948718,
+      "grad_norm": 0.0009031430818140507,
+      "learning_rate": 6.3390313390313394e-06,
+      "loss": 0.0001,
+      "step": 10050
+    },
+    {
+      "epoch": 42.00991452991453,
+      "grad_norm": 392.3472595214844,
+      "learning_rate": 6.334283000949668e-06,
+      "loss": 0.4313,
+      "step": 10060
+    },
+    {
+      "epoch": 42.01,
+      "eval_accuracy": 0.48,
+      "eval_loss": 4.089754581451416,
+      "eval_runtime": 33.5482,
+      "eval_samples_per_second": 0.745,
+      "eval_steps_per_second": 0.745,
+      "step": 10062
+    },
+    {
+      "epoch": 43.00034188034188,
+      "grad_norm": 0.004300788510590792,
+      "learning_rate": 6.329534662867997e-06,
+      "loss": 0.0001,
+      "step": 10070
+    },
+    {
+      "epoch": 43.00076923076923,
+      "grad_norm": 0.0026491975877434015,
+      "learning_rate": 6.324786324786325e-06,
+      "loss": 0.004,
+      "step": 10080
+    },
+    {
+      "epoch": 43.00119658119658,
+      "grad_norm": 0.002639173995703459,
+      "learning_rate": 6.3200379867046535e-06,
+      "loss": 0.0006,
+      "step": 10090
+    },
+    {
+      "epoch": 43.00162393162393,
+      "grad_norm": 0.0024889677297323942,
+      "learning_rate": 6.315289648622982e-06,
+      "loss": 0.0006,
+      "step": 10100
+    },
+    {
+      "epoch": 43.00205128205128,
+      "grad_norm": 0.0487215556204319,
+      "learning_rate": 6.310541310541312e-06,
+      "loss": 0.0001,
+      "step": 10110
+    },
+    {
+      "epoch": 43.002478632478635,
+      "grad_norm": 0.0023086927831172943,
+      "learning_rate": 6.30579297245964e-06,
+      "loss": 0.307,
+      "step": 10120
+    },
+    {
+      "epoch": 43.002905982905986,
+      "grad_norm": 0.007922650314867496,
+      "learning_rate": 6.301044634377968e-06,
+      "loss": 0.3353,
+      "step": 10130
+    },
+    {
+      "epoch": 43.00333333333333,
+      "grad_norm": 0.0018696904880926013,
+      "learning_rate": 6.296296296296297e-06,
+      "loss": 0.0131,
+      "step": 10140
+    },
+    {
+      "epoch": 43.00376068376068,
+      "grad_norm": 0.027236852794885635,
+      "learning_rate": 6.291547958214626e-06,
+      "loss": 0.017,
+      "step": 10150
+    },
+    {
+      "epoch": 43.00418803418803,
+      "grad_norm": 0.0007559856749139726,
+      "learning_rate": 6.286799620132954e-06,
+      "loss": 0.0005,
+      "step": 10160
+    },
+    {
+      "epoch": 43.004615384615384,
+      "grad_norm": 11.704588890075684,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 0.0014,
+      "step": 10170
+    },
+    {
+      "epoch": 43.005042735042736,
+      "grad_norm": 0.06166832521557808,
+      "learning_rate": 6.277302943969611e-06,
+      "loss": 0.9794,
+      "step": 10180
+    },
+    {
+      "epoch": 43.00547008547009,
+      "grad_norm": 0.0030286931432783604,
+      "learning_rate": 6.272554605887939e-06,
+      "loss": 0.2002,
+      "step": 10190
+    },
+    {
+      "epoch": 43.00589743589744,
+      "grad_norm": 0.006914252880960703,
+      "learning_rate": 6.267806267806268e-06,
+      "loss": 0.0001,
+      "step": 10200
+    },
+    {
+      "epoch": 43.00632478632479,
+      "grad_norm": 149.87620544433594,
+      "learning_rate": 6.2630579297245964e-06,
+      "loss": 1.023,
+      "step": 10210
+    },
+    {
+      "epoch": 43.006752136752134,
+      "grad_norm": 0.0074666948057711124,
+      "learning_rate": 6.2583095916429256e-06,
+      "loss": 0.2019,
+      "step": 10220
+    },
+    {
+      "epoch": 43.007179487179485,
+      "grad_norm": 0.0010838122107088566,
+      "learning_rate": 6.253561253561255e-06,
+      "loss": 0.0011,
+      "step": 10230
+    },
+    {
+      "epoch": 43.007606837606836,
+      "grad_norm": 0.10124364495277405,
+      "learning_rate": 6.248812915479583e-06,
+      "loss": 0.3781,
+      "step": 10240
+    },
+    {
+      "epoch": 43.00803418803419,
+      "grad_norm": 0.0029677078127861023,
+      "learning_rate": 6.244064577397911e-06,
+      "loss": 0.1722,
+      "step": 10250
+    },
+    {
+      "epoch": 43.00846153846154,
+      "grad_norm": 0.012870069593191147,
+      "learning_rate": 6.23931623931624e-06,
+      "loss": 0.0001,
+      "step": 10260
+    },
+    {
+      "epoch": 43.00888888888889,
+      "grad_norm": 0.0010046313982456923,
+      "learning_rate": 6.234567901234569e-06,
+      "loss": 0.0003,
+      "step": 10270
+    },
+    {
+      "epoch": 43.00931623931624,
+      "grad_norm": 0.0031989836134016514,
+      "learning_rate": 6.229819563152897e-06,
+      "loss": 0.0002,
+      "step": 10280
+    },
+    {
+      "epoch": 43.00974358974359,
+      "grad_norm": 0.0007750399527139962,
+      "learning_rate": 6.225071225071225e-06,
+      "loss": 0.0002,
+      "step": 10290
+    },
+    {
+      "epoch": 43.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 4.78167200088501,
+      "eval_runtime": 33.6921,
+      "eval_samples_per_second": 0.742,
+      "eval_steps_per_second": 0.742,
+      "step": 10296
+    },
+    {
+      "epoch": 44.00017094017094,
+      "grad_norm": 0.0008096142555586994,
+      "learning_rate": 6.220322886989554e-06,
+      "loss": 0.7658,
+      "step": 10300
+    },
+    {
+      "epoch": 44.00059829059829,
+      "grad_norm": 0.001107421237975359,
+      "learning_rate": 6.215574548907882e-06,
+      "loss": 0.0011,
+      "step": 10310
+    },
+    {
+      "epoch": 44.00102564102564,
+      "grad_norm": 147.27357482910156,
+      "learning_rate": 6.210826210826212e-06,
+      "loss": 0.8891,
+      "step": 10320
+    },
+    {
+      "epoch": 44.00145299145299,
+      "grad_norm": 0.21983663737773895,
+      "learning_rate": 6.20607787274454e-06,
+      "loss": 0.0002,
+      "step": 10330
+    },
+    {
+      "epoch": 44.001880341880344,
+      "grad_norm": 0.0018023523734882474,
+      "learning_rate": 6.2013295346628685e-06,
+      "loss": 0.0806,
+      "step": 10340
+    },
+    {
+      "epoch": 44.002307692307696,
+      "grad_norm": 0.0013478569453582168,
+      "learning_rate": 6.196581196581198e-06,
+      "loss": 0.442,
+      "step": 10350
+    },
+    {
+      "epoch": 44.00273504273504,
+      "grad_norm": 0.0480712465941906,
+      "learning_rate": 6.191832858499526e-06,
+      "loss": 1.1859,
+      "step": 10360
+    },
+    {
+      "epoch": 44.00316239316239,
+      "grad_norm": 0.0018559067975729704,
+      "learning_rate": 6.187084520417854e-06,
+      "loss": 0.8634,
+      "step": 10370
+    },
+    {
+      "epoch": 44.00358974358974,
+      "grad_norm": 0.0008481431868858635,
+      "learning_rate": 6.1823361823361825e-06,
+      "loss": 0.8489,
+      "step": 10380
+    },
+    {
+      "epoch": 44.004017094017094,
+      "grad_norm": 0.003446524264290929,
+      "learning_rate": 6.177587844254511e-06,
+      "loss": 0.5956,
+      "step": 10390
+    },
+    {
+      "epoch": 44.004444444444445,
+      "grad_norm": 0.0011318651959300041,
+      "learning_rate": 6.17283950617284e-06,
+      "loss": 0.0008,
+      "step": 10400
+    },
+    {
+      "epoch": 44.0048717948718,
+      "grad_norm": 0.002299872925505042,
+      "learning_rate": 6.168091168091168e-06,
+      "loss": 0.0322,
+      "step": 10410
+    },
+    {
+      "epoch": 44.00529914529915,
+      "grad_norm": 0.002232542959973216,
+      "learning_rate": 6.1633428300094974e-06,
+      "loss": 1.0857,
+      "step": 10420
+    },
+    {
+      "epoch": 44.0057264957265,
+      "grad_norm": 0.1128619983792305,
+      "learning_rate": 6.1585944919278266e-06,
+      "loss": 0.0205,
+      "step": 10430
+    },
+    {
+      "epoch": 44.00615384615384,
+      "grad_norm": 0.12662559747695923,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.876,
+      "step": 10440
+    },
+    {
+      "epoch": 44.006581196581195,
+      "grad_norm": 0.009139850735664368,
+      "learning_rate": 6.149097815764483e-06,
+      "loss": 0.0007,
+      "step": 10450
+    },
+    {
+      "epoch": 44.007008547008546,
+      "grad_norm": 0.0021971724927425385,
+      "learning_rate": 6.1443494776828115e-06,
+      "loss": 0.0002,
+      "step": 10460
+    },
+    {
+      "epoch": 44.0074358974359,
+      "grad_norm": 0.00538423378020525,
+      "learning_rate": 6.13960113960114e-06,
+      "loss": 0.0001,
+      "step": 10470
+    },
+    {
+      "epoch": 44.00786324786325,
+      "grad_norm": 0.026287034153938293,
+      "learning_rate": 6.134852801519469e-06,
+      "loss": 0.0432,
+      "step": 10480
+    },
+    {
+      "epoch": 44.0082905982906,
+      "grad_norm": 0.0007875883602537215,
+      "learning_rate": 6.130104463437797e-06,
+      "loss": 0.266,
+      "step": 10490
+    },
+    {
+      "epoch": 44.00871794871795,
+      "grad_norm": 0.01668292097747326,
+      "learning_rate": 6.1253561253561255e-06,
+      "loss": 0.0248,
+      "step": 10500
+    },
+    {
+      "epoch": 44.0091452991453,
+      "grad_norm": 0.025921987369656563,
+      "learning_rate": 6.120607787274454e-06,
+      "loss": 0.0002,
+      "step": 10510
+    },
+    {
+      "epoch": 44.00957264957265,
+      "grad_norm": 0.0005760848871432245,
+      "learning_rate": 6.115859449192783e-06,
+      "loss": 0.0901,
+      "step": 10520
+    },
+    {
+      "epoch": 44.01,
+      "grad_norm": 0.0007085147080942988,
+      "learning_rate": 6.111111111111112e-06,
+      "loss": 0.0001,
+      "step": 10530
+    },
+    {
+      "epoch": 44.01,
+      "eval_accuracy": 0.48,
+      "eval_loss": 4.866738319396973,
+      "eval_runtime": 33.6315,
+      "eval_samples_per_second": 0.743,
+      "eval_steps_per_second": 0.743,
+      "step": 10530
+    },
+    {
+      "epoch": 45.00042735042735,
+      "grad_norm": 0.013209497556090355,
+      "learning_rate": 6.10636277302944e-06,
+      "loss": 0.2852,
+      "step": 10540
+    },
+    {
+      "epoch": 45.0008547008547,
+      "grad_norm": 0.0015938766300678253,
+      "learning_rate": 6.101614434947769e-06,
+      "loss": 0.0001,
+      "step": 10550
+    },
+    {
+      "epoch": 45.001282051282054,
+      "grad_norm": 0.05549630522727966,
+      "learning_rate": 6.096866096866098e-06,
+      "loss": 0.9446,
+      "step": 10560
+    },
+    {
+      "epoch": 45.001709401709405,
+      "grad_norm": 0.0019539748318493366,
+      "learning_rate": 6.092117758784426e-06,
+      "loss": 0.0001,
+      "step": 10570
+    },
+    {
+      "epoch": 45.00213675213675,
+      "grad_norm": 0.0008734731236472726,
+      "learning_rate": 6.087369420702754e-06,
+      "loss": 0.0078,
+      "step": 10580
+    },
+    {
+      "epoch": 45.0025641025641,
+      "grad_norm": 0.0014504172140732408,
+      "learning_rate": 6.082621082621083e-06,
+      "loss": 0.0003,
+      "step": 10590
+    },
+    {
+      "epoch": 45.00299145299145,
+      "grad_norm": 0.0006624649395234883,
+      "learning_rate": 6.077872744539412e-06,
+      "loss": 0.0005,
+      "step": 10600
+    },
+    {
+      "epoch": 45.0034188034188,
+      "grad_norm": 0.0008454503258690238,
+      "learning_rate": 6.07312440645774e-06,
+      "loss": 0.7002,
+      "step": 10610
+    },
+    {
+      "epoch": 45.003846153846155,
+      "grad_norm": 0.0008091210620477796,
+      "learning_rate": 6.0683760683760684e-06,
+      "loss": 0.0006,
+      "step": 10620
+    },
+    {
+      "epoch": 45.004273504273506,
+      "grad_norm": 0.0016585165867581964,
+      "learning_rate": 6.063627730294398e-06,
+      "loss": 0.0001,
+      "step": 10630
+    },
+    {
+      "epoch": 45.00470085470086,
+      "grad_norm": 0.0022859878372401,
+      "learning_rate": 6.058879392212727e-06,
+      "loss": 0.1929,
+      "step": 10640
+    },
+    {
+      "epoch": 45.00512820512821,
+      "grad_norm": 604.373291015625,
+      "learning_rate": 6.054131054131055e-06,
+      "loss": 0.6576,
+      "step": 10650
+    },
+    {
+      "epoch": 45.00555555555555,
+      "grad_norm": 0.015343650244176388,
+      "learning_rate": 6.049382716049383e-06,
+      "loss": 0.555,
+      "step": 10660
+    },
+    {
+      "epoch": 45.005982905982904,
+      "grad_norm": 0.0035480374936014414,
+      "learning_rate": 6.044634377967712e-06,
+      "loss": 0.0039,
+      "step": 10670
+    },
+    {
+      "epoch": 45.006410256410255,
+      "grad_norm": 0.0072784130461514,
+      "learning_rate": 6.039886039886041e-06,
+      "loss": 0.0001,
+      "step": 10680
+    },
+    {
+      "epoch": 45.00683760683761,
+      "grad_norm": 0.0026009362190961838,
+      "learning_rate": 6.035137701804369e-06,
+      "loss": 0.0002,
+      "step": 10690
+    },
+    {
+      "epoch": 45.00726495726496,
+      "grad_norm": 0.000990421511232853,
+      "learning_rate": 6.030389363722697e-06,
+      "loss": 0.0012,
+      "step": 10700
+    },
+    {
+      "epoch": 45.00769230769231,
+      "grad_norm": 0.0006937840371392667,
+      "learning_rate": 6.025641025641026e-06,
+      "loss": 0.0,
+      "step": 10710
+    },
+    {
+      "epoch": 45.00811965811966,
+      "grad_norm": 0.001313463319092989,
+      "learning_rate": 6.020892687559354e-06,
+      "loss": 0.0073,
+      "step": 10720
+    },
+    {
+      "epoch": 45.00854700854701,
+      "grad_norm": 0.02181072160601616,
+      "learning_rate": 6.016144349477683e-06,
+      "loss": 0.0001,
+      "step": 10730
+    },
+    {
+      "epoch": 45.008974358974356,
+      "grad_norm": 0.2195536494255066,
+      "learning_rate": 6.011396011396012e-06,
+      "loss": 0.0003,
+      "step": 10740
+    },
+    {
+      "epoch": 45.00940170940171,
+      "grad_norm": 0.006073196418583393,
+      "learning_rate": 6.0066476733143405e-06,
+      "loss": 0.0004,
+      "step": 10750
+    },
+    {
+      "epoch": 45.00982905982906,
+      "grad_norm": 0.0016091590514406562,
+      "learning_rate": 6.00189933523267e-06,
+      "loss": 0.0007,
+      "step": 10760
+    },
+    {
+      "epoch": 45.01,
+      "eval_accuracy": 0.48,
+      "eval_loss": 4.561890125274658,
+      "eval_runtime": 34.2721,
+      "eval_samples_per_second": 0.729,
+      "eval_steps_per_second": 0.729,
+      "step": 10764
+    },
+    {
+      "epoch": 46.00025641025641,
+      "grad_norm": 0.0011347552062943578,
+      "learning_rate": 5.997150997150998e-06,
+      "loss": 0.7172,
+      "step": 10770
+    },
+    {
+      "epoch": 46.00068376068376,
+      "grad_norm": 375.4585266113281,
+      "learning_rate": 5.992402659069326e-06,
+      "loss": 0.821,
+      "step": 10780
+    },
+    {
+      "epoch": 46.00111111111111,
+      "grad_norm": 0.0010995555203408003,
+      "learning_rate": 5.9876543209876546e-06,
+      "loss": 0.1693,
+      "step": 10790
+    },
+    {
+      "epoch": 46.00153846153846,
+      "grad_norm": 0.11372661590576172,
+      "learning_rate": 5.982905982905983e-06,
+      "loss": 0.0002,
+      "step": 10800
+    },
+    {
+      "epoch": 46.00196581196581,
+      "grad_norm": 0.0009383049909956753,
+      "learning_rate": 5.978157644824312e-06,
+      "loss": 0.0001,
+      "step": 10810
+    },
+    {
+      "epoch": 46.00239316239316,
+      "grad_norm": 0.004879310727119446,
+      "learning_rate": 5.97340930674264e-06,
+      "loss": 0.0005,
+      "step": 10820
+    },
+    {
+      "epoch": 46.00282051282051,
+      "grad_norm": 0.00243366789072752,
+      "learning_rate": 5.968660968660969e-06,
+      "loss": 0.1072,
+      "step": 10830
+    },
+    {
+      "epoch": 46.003247863247864,
+      "grad_norm": 0.0007890159031376243,
+      "learning_rate": 5.963912630579299e-06,
+      "loss": 0.0003,
+      "step": 10840
+    },
+    {
+      "epoch": 46.003675213675216,
+      "grad_norm": 0.009691433981060982,
+      "learning_rate": 5.959164292497627e-06,
+      "loss": 0.4108,
+      "step": 10850
+    },
+    {
+      "epoch": 46.00410256410257,
+      "grad_norm": 0.003631254890933633,
+      "learning_rate": 5.954415954415955e-06,
+      "loss": 0.0001,
+      "step": 10860
+    },
+    {
+      "epoch": 46.00452991452991,
+      "grad_norm": 0.0005667143850587308,
+      "learning_rate": 5.9496676163342835e-06,
+      "loss": 1.0405,
+      "step": 10870
+    },
+    {
+      "epoch": 46.00495726495726,
+      "grad_norm": 0.0033552220556885004,
+      "learning_rate": 5.944919278252612e-06,
+      "loss": 0.0001,
+      "step": 10880
+    },
+    {
+      "epoch": 46.005384615384614,
+      "grad_norm": 0.07237747311592102,
+      "learning_rate": 5.940170940170941e-06,
+      "loss": 0.6576,
+      "step": 10890
+    },
+    {
+      "epoch": 46.005811965811965,
+      "grad_norm": 0.0010696067474782467,
+      "learning_rate": 5.935422602089269e-06,
+      "loss": 0.0001,
+      "step": 10900
+    },
+    {
+      "epoch": 46.006239316239316,
+      "grad_norm": 0.0015365901635959744,
+      "learning_rate": 5.9306742640075975e-06,
+      "loss": 0.7127,
+      "step": 10910
+    },
+    {
+      "epoch": 46.00666666666667,
+      "grad_norm": 0.002339027589187026,
+      "learning_rate": 5.925925925925926e-06,
+      "loss": 0.0023,
+      "step": 10920
+    },
+    {
+      "epoch": 46.00709401709402,
+      "grad_norm": 0.01849093660712242,
+      "learning_rate": 5.921177587844255e-06,
+      "loss": 0.4024,
+      "step": 10930
+    },
+    {
+      "epoch": 46.00752136752137,
+      "grad_norm": 0.009332343935966492,
+      "learning_rate": 5.916429249762583e-06,
+      "loss": 0.0002,
+      "step": 10940
+    },
+    {
+      "epoch": 46.007948717948715,
+      "grad_norm": 0.009404930286109447,
+      "learning_rate": 5.911680911680912e-06,
+      "loss": 0.001,
+      "step": 10950
+    },
+    {
+      "epoch": 46.008376068376066,
+      "grad_norm": 0.0012134173884987831,
+      "learning_rate": 5.906932573599241e-06,
+      "loss": 0.5511,
+      "step": 10960
+    },
+    {
+      "epoch": 46.00880341880342,
+      "grad_norm": 0.01179675292223692,
+      "learning_rate": 5.90218423551757e-06,
+      "loss": 0.0018,
+      "step": 10970
+    },
+    {
+      "epoch": 46.00923076923077,
+      "grad_norm": 0.0012383083812892437,
+      "learning_rate": 5.897435897435898e-06,
+      "loss": 0.7245,
+      "step": 10980
+    },
+    {
+      "epoch": 46.00965811965812,
+      "grad_norm": 0.001664446317590773,
+      "learning_rate": 5.8926875593542264e-06,
+      "loss": 0.0009,
+      "step": 10990
+    },
+    {
+      "epoch": 46.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 5.025023937225342,
+      "eval_runtime": 33.6102,
+      "eval_samples_per_second": 0.744,
+      "eval_steps_per_second": 0.744,
+      "step": 10998
+    },
+    {
+      "epoch": 47.00008547008547,
+      "grad_norm": 0.002813153900206089,
+      "learning_rate": 5.887939221272555e-06,
+      "loss": 0.6121,
+      "step": 11000
+    },
+    {
+      "epoch": 47.00051282051282,
+      "grad_norm": 0.002290428848937154,
+      "learning_rate": 5.883190883190884e-06,
+      "loss": 0.0002,
+      "step": 11010
+    },
+    {
+      "epoch": 47.00094017094017,
+      "grad_norm": 0.013784612528979778,
+      "learning_rate": 5.878442545109212e-06,
+      "loss": 0.0041,
+      "step": 11020
+    },
+    {
+      "epoch": 47.00136752136752,
+      "grad_norm": 0.008057618513703346,
+      "learning_rate": 5.8736942070275405e-06,
+      "loss": 0.0002,
+      "step": 11030
+    },
+    {
+      "epoch": 47.00179487179487,
+      "grad_norm": 0.0026002710219472647,
+      "learning_rate": 5.868945868945869e-06,
+      "loss": 0.9305,
+      "step": 11040
+    },
+    {
+      "epoch": 47.00222222222222,
+      "grad_norm": 375.0780029296875,
+      "learning_rate": 5.864197530864199e-06,
+      "loss": 0.9073,
+      "step": 11050
+    },
+    {
+      "epoch": 47.002649572649574,
+      "grad_norm": 0.002601037034764886,
+      "learning_rate": 5.859449192782527e-06,
+      "loss": 0.1539,
+      "step": 11060
+    },
+    {
+      "epoch": 47.003076923076925,
+      "grad_norm": 0.013979580253362656,
+      "learning_rate": 5.854700854700855e-06,
+      "loss": 0.0031,
+      "step": 11070
+    },
+    {
+      "epoch": 47.00350427350428,
+      "grad_norm": 0.003610535990446806,
+      "learning_rate": 5.849952516619184e-06,
+      "loss": 0.0126,
+      "step": 11080
+    },
+    {
+      "epoch": 47.00393162393162,
+      "grad_norm": 0.0018818234093487263,
+      "learning_rate": 5.845204178537513e-06,
+      "loss": 0.0001,
+      "step": 11090
+    },
+    {
+      "epoch": 47.00435897435897,
+      "grad_norm": 209.32333374023438,
+      "learning_rate": 5.840455840455841e-06,
+      "loss": 1.5139,
+      "step": 11100
+    },
+    {
+      "epoch": 47.00478632478632,
+      "grad_norm": 0.002293006982654333,
+      "learning_rate": 5.835707502374169e-06,
+      "loss": 0.0011,
+      "step": 11110
+    },
+    {
+      "epoch": 47.005213675213675,
+      "grad_norm": 0.0014583432348445058,
+      "learning_rate": 5.830959164292498e-06,
+      "loss": 0.0002,
+      "step": 11120
+    },
+    {
+      "epoch": 47.005641025641026,
+      "grad_norm": 0.000818538770545274,
+      "learning_rate": 5.826210826210826e-06,
+      "loss": 0.0164,
+      "step": 11130
+    },
+    {
+      "epoch": 47.00606837606838,
+      "grad_norm": 0.025977041572332382,
+      "learning_rate": 5.821462488129155e-06,
+      "loss": 0.0002,
+      "step": 11140
+    },
+    {
+      "epoch": 47.00649572649573,
+      "grad_norm": 0.0008319366024807096,
+      "learning_rate": 5.816714150047484e-06,
+      "loss": 0.0003,
+      "step": 11150
+    },
+    {
+      "epoch": 47.00692307692308,
+      "grad_norm": 0.007519662380218506,
+      "learning_rate": 5.8119658119658126e-06,
+      "loss": 0.0643,
+      "step": 11160
+    },
+    {
+      "epoch": 47.007350427350424,
+      "grad_norm": 0.14387542009353638,
+      "learning_rate": 5.807217473884142e-06,
+      "loss": 0.0002,
+      "step": 11170
+    },
+    {
+      "epoch": 47.007777777777775,
+      "grad_norm": 0.0030812029726803303,
+      "learning_rate": 5.80246913580247e-06,
+      "loss": 0.0001,
+      "step": 11180
+    },
+    {
+      "epoch": 47.00820512820513,
+      "grad_norm": 0.0011086307931691408,
+      "learning_rate": 5.797720797720798e-06,
+      "loss": 0.3969,
+      "step": 11190
+    },
+    {
+      "epoch": 47.00863247863248,
+      "grad_norm": 0.0029646113980561495,
+      "learning_rate": 5.792972459639127e-06,
+      "loss": 0.0001,
+      "step": 11200
+    },
+    {
+      "epoch": 47.00905982905983,
+      "grad_norm": 0.003931617829948664,
+      "learning_rate": 5.788224121557455e-06,
+      "loss": 0.7591,
+      "step": 11210
+    },
+    {
+      "epoch": 47.00948717948718,
+      "grad_norm": 0.006124768406152725,
+      "learning_rate": 5.783475783475784e-06,
+      "loss": 0.2924,
+      "step": 11220
+    },
+    {
+      "epoch": 47.00991452991453,
+      "grad_norm": 0.0010141300735995173,
+      "learning_rate": 5.778727445394112e-06,
+      "loss": 0.0001,
+      "step": 11230
+    },
+    {
+      "epoch": 47.01,
+      "eval_accuracy": 0.48,
+      "eval_loss": 4.412927627563477,
+      "eval_runtime": 33.7429,
+      "eval_samples_per_second": 0.741,
+      "eval_steps_per_second": 0.741,
+      "step": 11232
+    },
+    {
+      "epoch": 48.00034188034188,
+      "grad_norm": 0.007142414338886738,
+      "learning_rate": 5.773979107312441e-06,
+      "loss": 0.8418,
+      "step": 11240
+    },
+    {
+      "epoch": 48.00076923076923,
+      "grad_norm": 1.3151068687438965,
+      "learning_rate": 5.769230769230769e-06,
+      "loss": 1.6759,
+      "step": 11250
+    },
+    {
+      "epoch": 48.00119658119658,
+      "grad_norm": 0.0036189991515129805,
+      "learning_rate": 5.764482431149099e-06,
+      "loss": 0.0004,
+      "step": 11260
+    },
+    {
+      "epoch": 48.00162393162393,
+      "grad_norm": 0.000717452319804579,
+      "learning_rate": 5.759734093067427e-06,
+      "loss": 0.0001,
+      "step": 11270
+    },
+    {
+      "epoch": 48.00205128205128,
+      "grad_norm": 0.07057972997426987,
+      "learning_rate": 5.7549857549857555e-06,
+      "loss": 0.0422,
+      "step": 11280
+    },
+    {
+      "epoch": 48.002478632478635,
+      "grad_norm": 589.4522094726562,
+      "learning_rate": 5.750237416904084e-06,
+      "loss": 0.0541,
+      "step": 11290
+    },
+    {
+      "epoch": 48.002905982905986,
+      "grad_norm": 0.01362372562289238,
+      "learning_rate": 5.745489078822413e-06,
+      "loss": 0.0055,
+      "step": 11300
+    },
+    {
+      "epoch": 48.00333333333333,
+      "grad_norm": 0.01434069313108921,
+      "learning_rate": 5.740740740740741e-06,
+      "loss": 0.7408,
+      "step": 11310
+    },
+    {
+      "epoch": 48.00376068376068,
+      "grad_norm": 0.0020310671534389257,
+      "learning_rate": 5.7359924026590695e-06,
+      "loss": 0.0002,
+      "step": 11320
+    },
+    {
+      "epoch": 48.00418803418803,
+      "grad_norm": 0.005610823631286621,
+      "learning_rate": 5.731244064577398e-06,
+      "loss": 0.0002,
+      "step": 11330
+    },
+    {
+      "epoch": 48.004615384615384,
+      "grad_norm": 0.0008401147206313908,
+      "learning_rate": 5.726495726495727e-06,
+      "loss": 0.0001,
+      "step": 11340
+    },
+    {
+      "epoch": 48.005042735042736,
+      "grad_norm": 0.03286376968026161,
+      "learning_rate": 5.721747388414055e-06,
+      "loss": 0.0007,
+      "step": 11350
+    },
+    {
+      "epoch": 48.00547008547009,
+      "grad_norm": 0.002525913529098034,
+      "learning_rate": 5.716999050332384e-06,
+      "loss": 0.0003,
+      "step": 11360
+    },
+    {
+      "epoch": 48.00589743589744,
+      "grad_norm": 0.0011132374638691545,
+      "learning_rate": 5.7122507122507136e-06,
+      "loss": 0.0,
+      "step": 11370
+    },
+    {
+      "epoch": 48.00632478632479,
+      "grad_norm": 0.0005615410045720637,
+      "learning_rate": 5.707502374169042e-06,
+      "loss": 0.0002,
+      "step": 11380
+    },
+    {
+      "epoch": 48.006752136752134,
+      "grad_norm": 0.006012198980897665,
+      "learning_rate": 5.70275403608737e-06,
+      "loss": 0.0007,
+      "step": 11390
+    },
+    {
+      "epoch": 48.007179487179485,
+      "grad_norm": 0.014253268949687481,
+      "learning_rate": 5.6980056980056985e-06,
+      "loss": 0.0001,
+      "step": 11400
+    },
+    {
+      "epoch": 48.007606837606836,
+      "grad_norm": 0.001018693670630455,
+      "learning_rate": 5.693257359924027e-06,
+      "loss": 0.0002,
+      "step": 11410
+    },
+    {
+      "epoch": 48.00803418803419,
+      "grad_norm": 0.030744800344109535,
+      "learning_rate": 5.688509021842356e-06,
+      "loss": 0.6403,
+      "step": 11420
+    },
+    {
+      "epoch": 48.00846153846154,
+      "grad_norm": 0.013697023503482342,
+      "learning_rate": 5.683760683760684e-06,
+      "loss": 0.0001,
+      "step": 11430
+    },
+    {
+      "epoch": 48.00888888888889,
+      "grad_norm": 0.01648533344268799,
+      "learning_rate": 5.6790123456790125e-06,
+      "loss": 0.0002,
+      "step": 11440
+    },
+    {
+      "epoch": 48.00931623931624,
+      "grad_norm": 0.0017602647421881557,
+      "learning_rate": 5.674264007597341e-06,
+      "loss": 0.0014,
+      "step": 11450
+    },
+    {
+      "epoch": 48.00974358974359,
+      "grad_norm": 0.0007377556175924838,
+      "learning_rate": 5.669515669515669e-06,
+      "loss": 0.0001,
+      "step": 11460
+    },
+    {
+      "epoch": 48.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 5.5987443923950195,
+      "eval_runtime": 33.7645,
+      "eval_samples_per_second": 0.74,
+      "eval_steps_per_second": 0.74,
+      "step": 11466
+    },
+    {
+      "epoch": 49.00017094017094,
+      "grad_norm": 0.0007492690929211676,
+      "learning_rate": 5.664767331433999e-06,
+      "loss": 0.6818,
+      "step": 11470
+    },
+    {
+      "epoch": 49.00059829059829,
+      "grad_norm": 0.03045983612537384,
+      "learning_rate": 5.660018993352327e-06,
+      "loss": 0.5287,
+      "step": 11480
+    },
+    {
+      "epoch": 49.00102564102564,
+      "grad_norm": 0.0005993829108774662,
+      "learning_rate": 5.655270655270656e-06,
+      "loss": 0.0425,
+      "step": 11490
+    },
+    {
+      "epoch": 49.00145299145299,
+      "grad_norm": 0.0005655125132761896,
+      "learning_rate": 5.650522317188985e-06,
+      "loss": 0.0001,
+      "step": 11500
+    },
+    {
+      "epoch": 49.001880341880344,
+      "grad_norm": 0.000584998691920191,
+      "learning_rate": 5.645773979107313e-06,
+      "loss": 0.0001,
+      "step": 11510
+    },
+    {
+      "epoch": 49.002307692307696,
+      "grad_norm": 0.0036167309153825045,
+      "learning_rate": 5.641025641025641e-06,
+      "loss": 0.0019,
+      "step": 11520
+    },
+    {
+      "epoch": 49.00273504273504,
+      "grad_norm": 0.0014566316967830062,
+      "learning_rate": 5.63627730294397e-06,
+      "loss": 0.0002,
+      "step": 11530
+    },
+    {
+      "epoch": 49.00316239316239,
+      "grad_norm": 194.14332580566406,
+      "learning_rate": 5.631528964862298e-06,
+      "loss": 0.8389,
+      "step": 11540
+    },
+    {
+      "epoch": 49.00358974358974,
+      "grad_norm": 0.03927035257220268,
+      "learning_rate": 5.626780626780627e-06,
+      "loss": 0.0001,
+      "step": 11550
+    },
+    {
+      "epoch": 49.004017094017094,
+      "grad_norm": 0.01512466836720705,
+      "learning_rate": 5.6220322886989554e-06,
+      "loss": 0.0011,
+      "step": 11560
+    },
+    {
+      "epoch": 49.004444444444445,
+      "grad_norm": 0.0020757820457220078,
+      "learning_rate": 5.617283950617285e-06,
+      "loss": 0.3843,
+      "step": 11570
+    },
+    {
+      "epoch": 49.0048717948718,
+      "grad_norm": 0.0007370910025201738,
+      "learning_rate": 5.612535612535614e-06,
+      "loss": 0.0001,
+      "step": 11580
+    },
+    {
+      "epoch": 49.00529914529915,
+      "grad_norm": 0.0047119674272835255,
+      "learning_rate": 5.607787274453942e-06,
+      "loss": 0.4626,
+      "step": 11590
+    },
+    {
+      "epoch": 49.0057264957265,
+      "grad_norm": 0.0026911916211247444,
+      "learning_rate": 5.60303893637227e-06,
+      "loss": 0.0649,
+      "step": 11600
+    },
+    {
+      "epoch": 49.00615384615384,
+      "grad_norm": 0.02297067642211914,
+      "learning_rate": 5.598290598290599e-06,
+      "loss": 0.0036,
+      "step": 11610
+    },
+    {
+      "epoch": 49.006581196581195,
+      "grad_norm": 0.0033419374376535416,
+      "learning_rate": 5.593542260208927e-06,
+      "loss": 0.0001,
+      "step": 11620
+    },
+    {
+      "epoch": 49.007008547008546,
+      "grad_norm": 0.08777555078268051,
+      "learning_rate": 5.588793922127256e-06,
+      "loss": 0.95,
+      "step": 11630
+    },
+    {
+      "epoch": 49.0074358974359,
+      "grad_norm": 0.0006313354824669659,
+      "learning_rate": 5.584045584045584e-06,
+      "loss": 0.6174,
+      "step": 11640
+    },
+    {
+      "epoch": 49.00786324786325,
+      "grad_norm": 0.0009615861345082521,
+      "learning_rate": 5.579297245963913e-06,
+      "loss": 0.0004,
+      "step": 11650
+    },
+    {
+      "epoch": 49.0082905982906,
+      "grad_norm": 0.04878242313861847,
+      "learning_rate": 5.574548907882241e-06,
+      "loss": 0.0001,
+      "step": 11660
+    },
+    {
+      "epoch": 49.00871794871795,
+      "grad_norm": 0.0008145536412484944,
+      "learning_rate": 5.56980056980057e-06,
+      "loss": 0.0002,
+      "step": 11670
+    },
+    {
+      "epoch": 49.0091452991453,
+      "grad_norm": 0.0008402611128985882,
+      "learning_rate": 5.565052231718899e-06,
+      "loss": 0.0,
+      "step": 11680
+    },
+    {
+      "epoch": 49.00957264957265,
+      "grad_norm": 0.17107157409191132,
+      "learning_rate": 5.5603038936372275e-06,
+      "loss": 0.044,
+      "step": 11690
+    },
+    {
+      "epoch": 49.01,
+      "grad_norm": 0.007154208142310381,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 0.0003,
+      "step": 11700
+    },
+    {
+      "epoch": 49.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 5.4567461013793945,
+      "eval_runtime": 33.7217,
+      "eval_samples_per_second": 0.741,
+      "eval_steps_per_second": 0.741,
+      "step": 11700
+    },
+    {
+      "epoch": 50.00042735042735,
+      "grad_norm": 0.0006979930330999196,
+      "learning_rate": 5.550807217473885e-06,
+      "loss": 0.0,
+      "step": 11710
+    },
+    {
+      "epoch": 50.0008547008547,
+      "grad_norm": 0.005530244205147028,
+      "learning_rate": 5.546058879392213e-06,
+      "loss": 0.0002,
+      "step": 11720
+    },
+    {
+      "epoch": 50.001282051282054,
+      "grad_norm": 0.007461980450898409,
+      "learning_rate": 5.5413105413105416e-06,
+      "loss": 0.0723,
+      "step": 11730
+    },
+    {
+      "epoch": 50.001709401709405,
+      "grad_norm": 0.000708905456122011,
+      "learning_rate": 5.53656220322887e-06,
+      "loss": 0.0008,
+      "step": 11740
+    },
+    {
+      "epoch": 50.00213675213675,
+      "grad_norm": 0.0005582983139902353,
+      "learning_rate": 5.531813865147199e-06,
+      "loss": 0.2155,
+      "step": 11750
+    },
+    {
+      "epoch": 50.0025641025641,
+      "grad_norm": 0.001401066081598401,
+      "learning_rate": 5.527065527065527e-06,
+      "loss": 0.0042,
+      "step": 11760
+    },
+    {
+      "epoch": 50.00299145299145,
+      "grad_norm": 0.0036240883637219667,
+      "learning_rate": 5.522317188983856e-06,
+      "loss": 0.5656,
+      "step": 11770
+    },
+    {
+      "epoch": 50.0034188034188,
+      "grad_norm": 0.0006863917224109173,
+      "learning_rate": 5.517568850902186e-06,
+      "loss": 0.0002,
+      "step": 11780
+    },
+    {
+      "epoch": 50.003846153846155,
+      "grad_norm": 0.0006204345263540745,
+      "learning_rate": 5.512820512820514e-06,
+      "loss": 0.0001,
+      "step": 11790
+    },
+    {
+      "epoch": 50.004273504273506,
+      "grad_norm": 0.0025562657974660397,
+      "learning_rate": 5.508072174738842e-06,
+      "loss": 0.0004,
+      "step": 11800
+    },
+    {
+      "epoch": 50.00470085470086,
+      "grad_norm": 0.0017411591252312064,
+      "learning_rate": 5.5033238366571705e-06,
+      "loss": 0.0002,
+      "step": 11810
+    },
+    {
+      "epoch": 50.00512820512821,
+      "grad_norm": 23.52789306640625,
+      "learning_rate": 5.498575498575499e-06,
+      "loss": 0.0036,
+      "step": 11820
+    },
+    {
+      "epoch": 50.00555555555555,
+      "grad_norm": 0.0005795454489998519,
+      "learning_rate": 5.493827160493828e-06,
+      "loss": 0.7016,
+      "step": 11830
+    },
+    {
+      "epoch": 50.005982905982904,
+      "grad_norm": 0.0019489077385514975,
+      "learning_rate": 5.489078822412156e-06,
+      "loss": 0.0001,
+      "step": 11840
+    },
+    {
+      "epoch": 50.006410256410255,
+      "grad_norm": 0.0617222860455513,
+      "learning_rate": 5.4843304843304845e-06,
+      "loss": 0.0001,
+      "step": 11850
+    },
+    {
+      "epoch": 50.00683760683761,
+      "grad_norm": 403.005859375,
+      "learning_rate": 5.479582146248813e-06,
+      "loss": 0.1644,
+      "step": 11860
+    },
+    {
+      "epoch": 50.00726495726496,
+      "grad_norm": 0.14567570388317108,
+      "learning_rate": 5.474833808167141e-06,
+      "loss": 0.0004,
+      "step": 11870
+    },
+    {
+      "epoch": 50.00769230769231,
+      "grad_norm": 0.0005991229554638267,
+      "learning_rate": 5.470085470085471e-06,
+      "loss": 0.0898,
+      "step": 11880
+    },
+    {
+      "epoch": 50.00811965811966,
+      "grad_norm": 0.14940251410007477,
+      "learning_rate": 5.465337132003799e-06,
+      "loss": 0.0001,
+      "step": 11890
+    },
+    {
+      "epoch": 50.00854700854701,
+      "grad_norm": 0.028690291568636894,
+      "learning_rate": 5.460588793922128e-06,
+      "loss": 0.0023,
+      "step": 11900
+    },
+    {
+      "epoch": 50.008974358974356,
+      "grad_norm": 0.21514463424682617,
+      "learning_rate": 5.455840455840457e-06,
+      "loss": 0.6174,
+      "step": 11910
+    },
+    {
+      "epoch": 50.00940170940171,
+      "grad_norm": 0.0008885301067493856,
+      "learning_rate": 5.451092117758785e-06,
+      "loss": 0.0001,
+      "step": 11920
+    },
+    {
+      "epoch": 50.00982905982906,
+      "grad_norm": 0.0035741357132792473,
+      "learning_rate": 5.446343779677113e-06,
+      "loss": 0.0468,
+      "step": 11930
+    },
+    {
+      "epoch": 50.01,
+      "eval_accuracy": 0.48,
+      "eval_loss": 5.021820545196533,
+      "eval_runtime": 33.527,
+      "eval_samples_per_second": 0.746,
+      "eval_steps_per_second": 0.746,
+      "step": 11934
+    },
+    {
+      "epoch": 51.00025641025641,
+      "grad_norm": 0.11837821453809738,
+      "learning_rate": 5.441595441595442e-06,
+      "loss": 0.0001,
+      "step": 11940
+    },
+    {
+      "epoch": 51.00068376068376,
+      "grad_norm": 32.09723663330078,
+      "learning_rate": 5.43684710351377e-06,
+      "loss": 0.0062,
+      "step": 11950
+    },
+    {
+      "epoch": 51.00111111111111,
+      "grad_norm": 0.0008311708807013929,
+      "learning_rate": 5.432098765432099e-06,
+      "loss": 0.4382,
+      "step": 11960
+    },
+    {
+      "epoch": 51.00153846153846,
+      "grad_norm": 0.00047426484525203705,
+      "learning_rate": 5.4273504273504275e-06,
+      "loss": 0.0019,
+      "step": 11970
+    },
+    {
+      "epoch": 51.00196581196581,
+      "grad_norm": 0.0012167677050456405,
+      "learning_rate": 5.422602089268756e-06,
+      "loss": 0.0002,
+      "step": 11980
+    },
+    {
+      "epoch": 51.00239316239316,
+      "grad_norm": 0.0028268792666494846,
+      "learning_rate": 5.417853751187086e-06,
+      "loss": 0.0001,
+      "step": 11990
+    },
+    {
+      "epoch": 51.00282051282051,
+      "grad_norm": 0.0009294459596276283,
+      "learning_rate": 5.413105413105414e-06,
+      "loss": 0.0001,
+      "step": 12000
+    },
+    {
+      "epoch": 51.003247863247864,
+      "grad_norm": 0.014754371717572212,
+      "learning_rate": 5.408357075023742e-06,
+      "loss": 0.0001,
+      "step": 12010
+    },
+    {
+      "epoch": 51.003675213675216,
+      "grad_norm": 0.027568388730287552,
+      "learning_rate": 5.403608736942071e-06,
+      "loss": 0.0002,
+      "step": 12020
+    },
+    {
+      "epoch": 51.00410256410257,
+      "grad_norm": 0.07785722613334656,
+      "learning_rate": 5.3988603988604e-06,
+      "loss": 0.0001,
+      "step": 12030
+    },
+    {
+      "epoch": 51.00452991452991,
+      "grad_norm": 0.0007142764516174793,
+      "learning_rate": 5.394112060778728e-06,
+      "loss": 0.0003,
+      "step": 12040
+    },
+    {
+      "epoch": 51.00495726495726,
+      "grad_norm": 0.002452197717502713,
+      "learning_rate": 5.389363722697056e-06,
+      "loss": 0.0003,
+      "step": 12050
+    },
+    {
+      "epoch": 51.005384615384614,
+      "grad_norm": 0.0013394391862675548,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 0.0001,
+      "step": 12060
+    },
+    {
+      "epoch": 51.005811965811965,
+      "grad_norm": 0.0005540793645195663,
+      "learning_rate": 5.379867046533713e-06,
+      "loss": 0.0,
+      "step": 12070
+    },
+    {
+      "epoch": 51.006239316239316,
+      "grad_norm": 0.004768889397382736,
+      "learning_rate": 5.375118708452042e-06,
+      "loss": 0.0002,
+      "step": 12080
+    },
+    {
+      "epoch": 51.00666666666667,
+      "grad_norm": 0.0035481529776006937,
+      "learning_rate": 5.370370370370371e-06,
+      "loss": 0.004,
+      "step": 12090
+    },
+    {
+      "epoch": 51.00709401709402,
+      "grad_norm": 0.0004956016782671213,
+      "learning_rate": 5.3656220322886995e-06,
+      "loss": 0.0001,
+      "step": 12100
+    },
+    {
+      "epoch": 51.00752136752137,
+      "grad_norm": 0.0010804428020492196,
+      "learning_rate": 5.360873694207029e-06,
+      "loss": 0.0001,
+      "step": 12110
+    },
+    {
+      "epoch": 51.007948717948715,
+      "grad_norm": 0.004500087816268206,
+      "learning_rate": 5.356125356125357e-06,
+      "loss": 0.031,
+      "step": 12120
+    },
+    {
+      "epoch": 51.008376068376066,
+      "grad_norm": 0.0010846515651792288,
+      "learning_rate": 5.351377018043685e-06,
+      "loss": 1.3764,
+      "step": 12130
+    },
+    {
+      "epoch": 51.00880341880342,
+      "grad_norm": 0.0024583388585597277,
+      "learning_rate": 5.346628679962014e-06,
+      "loss": 0.0,
+      "step": 12140
+    },
+    {
+      "epoch": 51.00923076923077,
+      "grad_norm": 0.000765733711887151,
+      "learning_rate": 5.341880341880342e-06,
+      "loss": 0.0134,
+      "step": 12150
+    },
+    {
+      "epoch": 51.00965811965812,
+      "grad_norm": 0.0004111463204026222,
+      "learning_rate": 5.337132003798671e-06,
+      "loss": 0.187,
+      "step": 12160
+    },
+    {
+      "epoch": 51.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 5.326884746551514,
+      "eval_runtime": 33.7105,
+      "eval_samples_per_second": 0.742,
+      "eval_steps_per_second": 0.742,
+      "step": 12168
+    },
+    {
+      "epoch": 52.00008547008547,
+      "grad_norm": 0.0006366129964590073,
+      "learning_rate": 5.332383665716999e-06,
+      "loss": 0.2729,
+      "step": 12170
+    },
+    {
+      "epoch": 52.00051282051282,
+      "grad_norm": 0.00045478667016141117,
+      "learning_rate": 5.327635327635328e-06,
+      "loss": 0.0002,
+      "step": 12180
+    },
+    {
+      "epoch": 52.00094017094017,
+      "grad_norm": 0.0012099344749003649,
+      "learning_rate": 5.322886989553656e-06,
+      "loss": 0.0017,
+      "step": 12190
+    },
+    {
+      "epoch": 52.00136752136752,
+      "grad_norm": 0.010483755730092525,
+      "learning_rate": 5.318138651471986e-06,
+      "loss": 0.0001,
+      "step": 12200
+    },
+    {
+      "epoch": 52.00179487179487,
+      "grad_norm": 0.001030059065669775,
+      "learning_rate": 5.313390313390314e-06,
+      "loss": 0.0001,
+      "step": 12210
+    },
+    {
+      "epoch": 52.00222222222222,
+      "grad_norm": 0.005780795589089394,
+      "learning_rate": 5.3086419753086425e-06,
+      "loss": 0.0001,
+      "step": 12220
+    },
+    {
+      "epoch": 52.002649572649574,
+      "grad_norm": 0.001182667212560773,
+      "learning_rate": 5.303893637226971e-06,
+      "loss": 0.7791,
+      "step": 12230
+    },
+    {
+      "epoch": 52.003076923076925,
+      "grad_norm": 0.10405033081769943,
+      "learning_rate": 5.2991452991453e-06,
+      "loss": 0.172,
+      "step": 12240
+    },
+    {
+      "epoch": 52.00350427350428,
+      "grad_norm": 0.0005188258364796638,
+      "learning_rate": 5.294396961063628e-06,
+      "loss": 0.0001,
+      "step": 12250
+    },
+    {
+      "epoch": 52.00393162393162,
+      "grad_norm": 0.0017536180093884468,
+      "learning_rate": 5.2896486229819565e-06,
+      "loss": 0.0001,
+      "step": 12260
+    },
+    {
+      "epoch": 52.00435897435897,
+      "grad_norm": 0.0009690941660664976,
+      "learning_rate": 5.284900284900285e-06,
+      "loss": 0.0001,
+      "step": 12270
+    },
+    {
+      "epoch": 52.00478632478632,
+      "grad_norm": 0.0005580552387982607,
+      "learning_rate": 5.280151946818613e-06,
+      "loss": 0.0003,
+      "step": 12280
+    },
+    {
+      "epoch": 52.005213675213675,
+      "grad_norm": 0.1803247481584549,
+      "learning_rate": 5.275403608736942e-06,
+      "loss": 0.0001,
+      "step": 12290
+    },
+    {
+      "epoch": 52.005641025641026,
+      "grad_norm": 0.005001471843570471,
+      "learning_rate": 5.270655270655271e-06,
+      "loss": 0.0002,
+      "step": 12300
+    },
+    {
+      "epoch": 52.00606837606838,
+      "grad_norm": 0.004487255588173866,
+      "learning_rate": 5.2659069325736e-06,
+      "loss": 0.5693,
+      "step": 12310
+    },
+    {
+      "epoch": 52.00649572649573,
+      "grad_norm": 612.653564453125,
+      "learning_rate": 5.261158594491929e-06,
+      "loss": 0.7203,
+      "step": 12320
+    },
+    {
+      "epoch": 52.00692307692308,
+      "grad_norm": 0.0006710452144034207,
+      "learning_rate": 5.256410256410257e-06,
+      "loss": 0.0004,
+      "step": 12330
+    },
+    {
+      "epoch": 52.007350427350424,
+      "grad_norm": 0.0005012887413613498,
+      "learning_rate": 5.2516619183285854e-06,
+      "loss": 0.5105,
+      "step": 12340
+    },
+    {
+      "epoch": 52.007777777777775,
+      "grad_norm": 0.002993144793435931,
+      "learning_rate": 5.246913580246914e-06,
+      "loss": 0.8932,
+      "step": 12350
+    },
+    {
+      "epoch": 52.00820512820513,
+      "grad_norm": 0.2561093270778656,
+      "learning_rate": 5.242165242165243e-06,
+      "loss": 0.0001,
+      "step": 12360
+    },
+    {
+      "epoch": 52.00863247863248,
+      "grad_norm": 0.003413307713344693,
+      "learning_rate": 5.237416904083571e-06,
+      "loss": 0.771,
+      "step": 12370
+    },
+    {
+      "epoch": 52.00905982905983,
+      "grad_norm": 0.000699891708791256,
+      "learning_rate": 5.2326685660018995e-06,
+      "loss": 0.8725,
+      "step": 12380
+    },
+    {
+      "epoch": 52.00948717948718,
+      "grad_norm": 0.0005549703491851687,
+      "learning_rate": 5.227920227920228e-06,
+      "loss": 0.0001,
+      "step": 12390
+    },
+    {
+      "epoch": 52.00991452991453,
+      "grad_norm": 0.02157350443303585,
+      "learning_rate": 5.223171889838556e-06,
+      "loss": 0.0002,
+      "step": 12400
+    },
+    {
+      "epoch": 52.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 5.436356067657471,
+      "eval_runtime": 33.5444,
+      "eval_samples_per_second": 0.745,
+      "eval_steps_per_second": 0.745,
+      "step": 12402
+    },
+    {
+      "epoch": 53.00034188034188,
+      "grad_norm": 0.0005383410025388002,
+      "learning_rate": 5.218423551756886e-06,
+      "loss": 0.0001,
+      "step": 12410
+    },
+    {
+      "epoch": 53.00076923076923,
+      "grad_norm": 0.0005392262828536332,
+      "learning_rate": 5.213675213675214e-06,
+      "loss": 0.0001,
+      "step": 12420
+    },
+    {
+      "epoch": 53.00119658119658,
+      "grad_norm": 3.4017317295074463,
+      "learning_rate": 5.208926875593543e-06,
+      "loss": 0.8486,
+      "step": 12430
+    },
+    {
+      "epoch": 53.00162393162393,
+      "grad_norm": 0.0005204902845434844,
+      "learning_rate": 5.204178537511872e-06,
+      "loss": 0.7513,
+      "step": 12440
+    },
+    {
+      "epoch": 53.00205128205128,
+      "grad_norm": 0.05535678565502167,
+      "learning_rate": 5.1994301994302e-06,
+      "loss": 0.0001,
+      "step": 12450
+    },
+    {
+      "epoch": 53.002478632478635,
+      "grad_norm": 0.002707291394472122,
+      "learning_rate": 5.194681861348528e-06,
+      "loss": 0.0001,
+      "step": 12460
+    },
+    {
+      "epoch": 53.002905982905986,
+      "grad_norm": 0.000867334078066051,
+      "learning_rate": 5.189933523266857e-06,
+      "loss": 0.0008,
+      "step": 12470
+    },
+    {
+      "epoch": 53.00333333333333,
+      "grad_norm": 0.0019842975307255983,
+      "learning_rate": 5.185185185185185e-06,
+      "loss": 0.0007,
+      "step": 12480
+    },
+    {
+      "epoch": 53.00376068376068,
+      "grad_norm": 0.002031872980296612,
+      "learning_rate": 5.180436847103514e-06,
+      "loss": 0.0013,
+      "step": 12490
+    },
+    {
+      "epoch": 53.00418803418803,
+      "grad_norm": 0.8920242786407471,
+      "learning_rate": 5.175688509021842e-06,
+      "loss": 0.0004,
+      "step": 12500
+    },
+    {
+      "epoch": 53.004615384615384,
+      "grad_norm": 0.005432880017906427,
+      "learning_rate": 5.1709401709401716e-06,
+      "loss": 0.0011,
+      "step": 12510
+    },
+    {
+      "epoch": 53.005042735042736,
+      "grad_norm": 0.0005356788169592619,
+      "learning_rate": 5.166191832858501e-06,
+      "loss": 0.1904,
+      "step": 12520
+    },
+    {
+      "epoch": 53.00547008547009,
+      "grad_norm": 0.00788265373557806,
+      "learning_rate": 5.161443494776829e-06,
+      "loss": 0.0093,
+      "step": 12530
+    },
+    {
+      "epoch": 53.00589743589744,
+      "grad_norm": 0.0006382990977726877,
+      "learning_rate": 5.156695156695157e-06,
+      "loss": 0.0003,
+      "step": 12540
+    },
+    {
+      "epoch": 53.00632478632479,
+      "grad_norm": 0.0006224510143510997,
+      "learning_rate": 5.151946818613486e-06,
+      "loss": 0.012,
+      "step": 12550
+    },
+    {
+      "epoch": 53.006752136752134,
+      "grad_norm": 0.00042305790702812374,
+      "learning_rate": 5.147198480531814e-06,
+      "loss": 0.0001,
+      "step": 12560
+    },
+    {
+      "epoch": 53.007179487179485,
+      "grad_norm": 0.00042074365774169564,
+      "learning_rate": 5.142450142450143e-06,
+      "loss": 0.0009,
+      "step": 12570
+    },
+    {
+      "epoch": 53.007606837606836,
+      "grad_norm": 0.00039612961700186133,
+      "learning_rate": 5.137701804368471e-06,
+      "loss": 0.0001,
+      "step": 12580
+    },
+    {
+      "epoch": 53.00803418803419,
+      "grad_norm": 0.029377451166510582,
+      "learning_rate": 5.1329534662868e-06,
+      "loss": 0.0,
+      "step": 12590
+    },
+    {
+      "epoch": 53.00846153846154,
+      "grad_norm": 0.0005169266369193792,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 0.0001,
+      "step": 12600
+    },
+    {
+      "epoch": 53.00888888888889,
+      "grad_norm": 0.007242798339575529,
+      "learning_rate": 5.123456790123458e-06,
+      "loss": 0.0,
+      "step": 12610
+    },
+    {
+      "epoch": 53.00931623931624,
+      "grad_norm": 0.01810172200202942,
+      "learning_rate": 5.118708452041786e-06,
+      "loss": 0.0055,
+      "step": 12620
+    },
+    {
+      "epoch": 53.00974358974359,
+      "grad_norm": 0.00594985531643033,
+      "learning_rate": 5.1139601139601145e-06,
+      "loss": 0.0001,
+      "step": 12630
+    },
+    {
+      "epoch": 53.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 5.730712890625,
+      "eval_runtime": 33.9799,
+      "eval_samples_per_second": 0.736,
+      "eval_steps_per_second": 0.736,
+      "step": 12636
+    },
+    {
+      "epoch": 54.00017094017094,
+      "grad_norm": 0.0023435179609805346,
+      "learning_rate": 5.109211775878443e-06,
+      "loss": 0.0001,
+      "step": 12640
+    },
+    {
+      "epoch": 54.00059829059829,
+      "grad_norm": 8.973128318786621,
+      "learning_rate": 5.104463437796772e-06,
+      "loss": 0.0012,
+      "step": 12650
+    },
+    {
+      "epoch": 54.00102564102564,
+      "grad_norm": 0.008565773256123066,
+      "learning_rate": 5.0997150997151e-06,
+      "loss": 0.0003,
+      "step": 12660
+    },
+    {
+      "epoch": 54.00145299145299,
+      "grad_norm": 0.0006859298446215689,
+      "learning_rate": 5.0949667616334285e-06,
+      "loss": 0.0,
+      "step": 12670
+    },
+    {
+      "epoch": 54.001880341880344,
+      "grad_norm": 0.0004111983289476484,
+      "learning_rate": 5.090218423551757e-06,
+      "loss": 0.0001,
+      "step": 12680
+    },
+    {
+      "epoch": 54.002307692307696,
+      "grad_norm": 0.0004719541466329247,
+      "learning_rate": 5.085470085470086e-06,
+      "loss": 0.5531,
+      "step": 12690
+    },
+    {
+      "epoch": 54.00273504273504,
+      "grad_norm": 0.004437173251062632,
+      "learning_rate": 5.080721747388414e-06,
+      "loss": 0.006,
+      "step": 12700
+    },
+    {
+      "epoch": 54.00316239316239,
+      "grad_norm": 0.0022185053676366806,
+      "learning_rate": 5.075973409306743e-06,
+      "loss": 0.0003,
+      "step": 12710
+    },
+    {
+      "epoch": 54.00358974358974,
+      "grad_norm": 0.0506693460047245,
+      "learning_rate": 5.071225071225072e-06,
+      "loss": 0.0001,
+      "step": 12720
+    },
+    {
+      "epoch": 54.004017094017094,
+      "grad_norm": 0.0014226617058739066,
+      "learning_rate": 5.066476733143401e-06,
+      "loss": 0.5765,
+      "step": 12730
+    },
+    {
+      "epoch": 54.004444444444445,
+      "grad_norm": 0.0005160128348506987,
+      "learning_rate": 5.061728395061729e-06,
+      "loss": 0.0,
+      "step": 12740
+    },
+    {
+      "epoch": 54.0048717948718,
+      "grad_norm": 0.0004296216939110309,
+      "learning_rate": 5.0569800569800575e-06,
+      "loss": 0.3456,
+      "step": 12750
+    },
+    {
+      "epoch": 54.00529914529915,
+      "grad_norm": 0.0012724578846246004,
+      "learning_rate": 5.052231718898386e-06,
+      "loss": 0.072,
+      "step": 12760
+    },
+    {
+      "epoch": 54.0057264957265,
+      "grad_norm": 0.0030480227433145046,
+      "learning_rate": 5.047483380816715e-06,
+      "loss": 0.0001,
+      "step": 12770
+    },
+    {
+      "epoch": 54.00615384615384,
+      "grad_norm": 0.0005519132246263325,
+      "learning_rate": 5.042735042735043e-06,
+      "loss": 0.0,
+      "step": 12780
+    },
+    {
+      "epoch": 54.006581196581195,
+      "grad_norm": 0.0007735707913525403,
+      "learning_rate": 5.0379867046533715e-06,
+      "loss": 0.4349,
+      "step": 12790
+    },
+    {
+      "epoch": 54.007008547008546,
+      "grad_norm": 0.0024715519975870848,
+      "learning_rate": 5.0332383665717e-06,
+      "loss": 0.9521,
+      "step": 12800
+    },
+    {
+      "epoch": 54.0074358974359,
+      "grad_norm": 0.005817775148898363,
+      "learning_rate": 5.028490028490028e-06,
+      "loss": 0.0001,
+      "step": 12810
+    },
+    {
+      "epoch": 54.00786324786325,
+      "grad_norm": 0.0017352089053019881,
+      "learning_rate": 5.023741690408358e-06,
+      "loss": 0.0004,
+      "step": 12820
+    },
+    {
+      "epoch": 54.0082905982906,
+      "grad_norm": 0.0010756135452538729,
+      "learning_rate": 5.018993352326686e-06,
+      "loss": 0.0001,
+      "step": 12830
+    },
+    {
+      "epoch": 54.00871794871795,
+      "grad_norm": 0.003498775651678443,
+      "learning_rate": 5.014245014245015e-06,
+      "loss": 0.0003,
+      "step": 12840
+    },
+    {
+      "epoch": 54.0091452991453,
+      "grad_norm": 0.010095912963151932,
+      "learning_rate": 5.009496676163344e-06,
+      "loss": 0.017,
+      "step": 12850
+    },
+    {
+      "epoch": 54.00957264957265,
+      "grad_norm": 0.00034551063436083496,
+      "learning_rate": 5.004748338081672e-06,
+      "loss": 0.0026,
+      "step": 12860
+    },
+    {
+      "epoch": 54.01,
+      "grad_norm": 0.0008016827632673085,
+      "learning_rate": 5e-06,
+      "loss": 0.0,
+      "step": 12870
+    },
+    {
+      "epoch": 54.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 5.978097915649414,
+      "eval_runtime": 33.5999,
+      "eval_samples_per_second": 0.744,
+      "eval_steps_per_second": 0.744,
+      "step": 12870
+    },
+    {
+      "epoch": 55.00042735042735,
+      "grad_norm": 0.0008040638058446348,
+      "learning_rate": 4.995251661918329e-06,
+      "loss": 0.0,
+      "step": 12880
+    },
+    {
+      "epoch": 55.0008547008547,
+      "grad_norm": 0.0010009427787736058,
+      "learning_rate": 4.990503323836657e-06,
+      "loss": 0.0243,
+      "step": 12890
+    },
+    {
+      "epoch": 55.001282051282054,
+      "grad_norm": 0.6322616934776306,
+      "learning_rate": 4.985754985754986e-06,
+      "loss": 0.0002,
+      "step": 12900
+    },
+    {
+      "epoch": 55.001709401709405,
+      "grad_norm": 0.00035148794995620847,
+      "learning_rate": 4.981006647673315e-06,
+      "loss": 0.0001,
+      "step": 12910
+    },
+    {
+      "epoch": 55.00213675213675,
+      "grad_norm": 0.0009499528096057475,
+      "learning_rate": 4.976258309591644e-06,
+      "loss": 0.158,
+      "step": 12920
+    },
+    {
+      "epoch": 55.0025641025641,
+      "grad_norm": 1.7123631238937378,
+      "learning_rate": 4.971509971509972e-06,
+      "loss": 0.0004,
+      "step": 12930
+    },
+    {
+      "epoch": 55.00299145299145,
+      "grad_norm": 0.00044161363621242344,
+      "learning_rate": 4.9667616334283e-06,
+      "loss": 0.0001,
+      "step": 12940
+    },
+    {
+      "epoch": 55.0034188034188,
+      "grad_norm": 0.0013492783764377236,
+      "learning_rate": 4.962013295346629e-06,
+      "loss": 0.4795,
+      "step": 12950
+    },
+    {
+      "epoch": 55.003846153846155,
+      "grad_norm": 0.0004561662208288908,
+      "learning_rate": 4.957264957264958e-06,
+      "loss": 0.0001,
+      "step": 12960
+    },
+    {
+      "epoch": 55.004273504273506,
+      "grad_norm": 0.00044428068213164806,
+      "learning_rate": 4.952516619183286e-06,
+      "loss": 0.001,
+      "step": 12970
+    },
+    {
+      "epoch": 55.00470085470086,
+      "grad_norm": 0.0006210142164491117,
+      "learning_rate": 4.947768281101615e-06,
+      "loss": 0.0001,
+      "step": 12980
+    },
+    {
+      "epoch": 55.00512820512821,
+      "grad_norm": 0.0037781489081680775,
+      "learning_rate": 4.943019943019943e-06,
+      "loss": 0.0,
+      "step": 12990
+    },
+    {
+      "epoch": 55.00555555555555,
+      "grad_norm": 0.0003753666242118925,
+      "learning_rate": 4.938271604938272e-06,
+      "loss": 0.0001,
+      "step": 13000
+    },
+    {
+      "epoch": 55.005982905982904,
+      "grad_norm": 0.009278559125959873,
+      "learning_rate": 4.933523266856601e-06,
+      "loss": 0.0001,
+      "step": 13010
+    },
+    {
+      "epoch": 55.006410256410255,
+      "grad_norm": 1.3523920774459839,
+      "learning_rate": 4.928774928774929e-06,
+      "loss": 0.0003,
+      "step": 13020
+    },
+    {
+      "epoch": 55.00683760683761,
+      "grad_norm": 0.00043029204243794084,
+      "learning_rate": 4.924026590693257e-06,
+      "loss": 0.0001,
+      "step": 13030
+    },
+    {
+      "epoch": 55.00726495726496,
+      "grad_norm": 0.00037467462243512273,
+      "learning_rate": 4.9192782526115865e-06,
+      "loss": 0.0001,
+      "step": 13040
+    },
+    {
+      "epoch": 55.00769230769231,
+      "grad_norm": 0.006767070386558771,
+      "learning_rate": 4.914529914529915e-06,
+      "loss": 0.0001,
+      "step": 13050
+    },
+    {
+      "epoch": 55.00811965811966,
+      "grad_norm": 0.00043661269592121243,
+      "learning_rate": 4.909781576448244e-06,
+      "loss": 0.0002,
+      "step": 13060
+    },
+    {
+      "epoch": 55.00854700854701,
+      "grad_norm": 0.0015735122142359614,
+      "learning_rate": 4.905033238366572e-06,
+      "loss": 0.0002,
+      "step": 13070
+    },
+    {
+      "epoch": 55.008974358974356,
+      "grad_norm": 0.009783231653273106,
+      "learning_rate": 4.9002849002849006e-06,
+      "loss": 0.0001,
+      "step": 13080
+    },
+    {
+      "epoch": 55.00940170940171,
+      "grad_norm": 0.000962352380156517,
+      "learning_rate": 4.895536562203229e-06,
+      "loss": 0.9947,
+      "step": 13090
+    },
+    {
+      "epoch": 55.00982905982906,
+      "grad_norm": 0.0005439156084321439,
+      "learning_rate": 4.890788224121558e-06,
+      "loss": 0.0001,
+      "step": 13100
+    },
+    {
+      "epoch": 55.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 4.822087287902832,
+      "eval_runtime": 33.7892,
+      "eval_samples_per_second": 0.74,
+      "eval_steps_per_second": 0.74,
+      "step": 13104
+    },
+    {
+      "epoch": 56.00025641025641,
+      "grad_norm": 0.010900448076426983,
+      "learning_rate": 4.886039886039886e-06,
+      "loss": 0.0001,
+      "step": 13110
+    },
+    {
+      "epoch": 56.00068376068376,
+      "grad_norm": 0.0009302808903157711,
+      "learning_rate": 4.8812915479582154e-06,
+      "loss": 0.0001,
+      "step": 13120
+    },
+    {
+      "epoch": 56.00111111111111,
+      "grad_norm": 0.0025970793794840574,
+      "learning_rate": 4.876543209876544e-06,
+      "loss": 0.582,
+      "step": 13130
+    },
+    {
+      "epoch": 56.00153846153846,
+      "grad_norm": 0.003855284536257386,
+      "learning_rate": 4.871794871794872e-06,
+      "loss": 0.0001,
+      "step": 13140
+    },
+    {
+      "epoch": 56.00196581196581,
+      "grad_norm": 0.16761374473571777,
+      "learning_rate": 4.8670465337132e-06,
+      "loss": 0.7305,
+      "step": 13150
+    },
+    {
+      "epoch": 56.00239316239316,
+      "grad_norm": 0.01091548427939415,
+      "learning_rate": 4.8622981956315295e-06,
+      "loss": 0.0001,
+      "step": 13160
+    },
+    {
+      "epoch": 56.00282051282051,
+      "grad_norm": 0.0012768160086125135,
+      "learning_rate": 4.857549857549858e-06,
+      "loss": 0.0184,
+      "step": 13170
+    },
+    {
+      "epoch": 56.003247863247864,
+      "grad_norm": 0.0003495585115160793,
+      "learning_rate": 4.852801519468187e-06,
+      "loss": 0.0,
+      "step": 13180
+    },
+    {
+      "epoch": 56.003675213675216,
+      "grad_norm": 0.00034958813921548426,
+      "learning_rate": 4.848053181386515e-06,
+      "loss": 0.19,
+      "step": 13190
+    },
+    {
+      "epoch": 56.00410256410257,
+      "grad_norm": 0.00039932169602252543,
+      "learning_rate": 4.8433048433048435e-06,
+      "loss": 0.0,
+      "step": 13200
+    },
+    {
+      "epoch": 56.00452991452991,
+      "grad_norm": 0.00140125781763345,
+      "learning_rate": 4.838556505223173e-06,
+      "loss": 0.4053,
+      "step": 13210
+    },
+    {
+      "epoch": 56.00495726495726,
+      "grad_norm": 0.00046302462578751147,
+      "learning_rate": 4.833808167141501e-06,
+      "loss": 0.1168,
+      "step": 13220
+    },
+    {
+      "epoch": 56.005384615384614,
+      "grad_norm": 0.07333427667617798,
+      "learning_rate": 4.829059829059829e-06,
+      "loss": 0.0,
+      "step": 13230
+    },
+    {
+      "epoch": 56.005811965811965,
+      "grad_norm": 0.5369700789451599,
+      "learning_rate": 4.824311490978158e-06,
+      "loss": 0.0002,
+      "step": 13240
+    },
+    {
+      "epoch": 56.006239316239316,
+      "grad_norm": 0.0011444406118243933,
+      "learning_rate": 4.819563152896487e-06,
+      "loss": 0.0001,
+      "step": 13250
+    },
+    {
+      "epoch": 56.00666666666667,
+      "grad_norm": 0.005999968387186527,
+      "learning_rate": 4.814814814814815e-06,
+      "loss": 0.0003,
+      "step": 13260
+    },
+    {
+      "epoch": 56.00709401709402,
+      "grad_norm": 0.006480031181126833,
+      "learning_rate": 4.810066476733144e-06,
+      "loss": 0.1278,
+      "step": 13270
+    },
+    {
+      "epoch": 56.00752136752137,
+      "grad_norm": 0.001407441683113575,
+      "learning_rate": 4.8053181386514724e-06,
+      "loss": 0.0011,
+      "step": 13280
+    },
+    {
+      "epoch": 56.007948717948715,
+      "grad_norm": 0.008889264427125454,
+      "learning_rate": 4.800569800569801e-06,
+      "loss": 0.0001,
+      "step": 13290
+    },
+    {
+      "epoch": 56.008376068376066,
+      "grad_norm": 0.0005853257607668638,
+      "learning_rate": 4.795821462488129e-06,
+      "loss": 0.0242,
+      "step": 13300
+    },
+    {
+      "epoch": 56.00880341880342,
+      "grad_norm": 0.0004666523600462824,
+      "learning_rate": 4.791073124406458e-06,
+      "loss": 0.0001,
+      "step": 13310
+    },
+    {
+      "epoch": 56.00923076923077,
+      "grad_norm": 404.7363586425781,
+      "learning_rate": 4.786324786324787e-06,
+      "loss": 0.7525,
+      "step": 13320
+    },
+    {
+      "epoch": 56.00965811965812,
+      "grad_norm": 0.0004823113849852234,
+      "learning_rate": 4.781576448243116e-06,
+      "loss": 0.0001,
+      "step": 13330
+    },
+    {
+      "epoch": 56.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 5.58083438873291,
+      "eval_runtime": 33.7533,
+      "eval_samples_per_second": 0.741,
+      "eval_steps_per_second": 0.741,
+      "step": 13338
+    },
+    {
+      "epoch": 57.00008547008547,
+      "grad_norm": 0.001120713190175593,
+      "learning_rate": 4.776828110161444e-06,
+      "loss": 0.0001,
+      "step": 13340
+    },
+    {
+      "epoch": 57.00051282051282,
+      "grad_norm": 0.0011573611991479993,
+      "learning_rate": 4.772079772079772e-06,
+      "loss": 0.0,
+      "step": 13350
+    },
+    {
+      "epoch": 57.00094017094017,
+      "grad_norm": 0.0003190473944414407,
+      "learning_rate": 4.7673314339981005e-06,
+      "loss": 0.0,
+      "step": 13360
+    },
+    {
+      "epoch": 57.00136752136752,
+      "grad_norm": 0.29808613657951355,
+      "learning_rate": 4.76258309591643e-06,
+      "loss": 0.0003,
+      "step": 13370
+    },
+    {
+      "epoch": 57.00179487179487,
+      "grad_norm": 0.0006945320055820048,
+      "learning_rate": 4.757834757834758e-06,
+      "loss": 0.0001,
+      "step": 13380
+    },
+    {
+      "epoch": 57.00222222222222,
+      "grad_norm": 0.0003567099920473993,
+      "learning_rate": 4.753086419753087e-06,
+      "loss": 0.0001,
+      "step": 13390
+    },
+    {
+      "epoch": 57.002649572649574,
+      "grad_norm": 0.012886996380984783,
+      "learning_rate": 4.748338081671415e-06,
+      "loss": 0.0,
+      "step": 13400
+    },
+    {
+      "epoch": 57.003076923076925,
+      "grad_norm": 0.0008997747790999711,
+      "learning_rate": 4.743589743589744e-06,
+      "loss": 0.0001,
+      "step": 13410
+    },
+    {
+      "epoch": 57.00350427350428,
+      "grad_norm": 0.008139757439494133,
+      "learning_rate": 4.738841405508073e-06,
+      "loss": 0.0,
+      "step": 13420
+    },
+    {
+      "epoch": 57.00393162393162,
+      "grad_norm": 0.008020474575459957,
+      "learning_rate": 4.734093067426401e-06,
+      "loss": 0.0001,
+      "step": 13430
+    },
+    {
+      "epoch": 57.00435897435897,
+      "grad_norm": 0.0004538062203209847,
+      "learning_rate": 4.729344729344729e-06,
+      "loss": 0.0001,
+      "step": 13440
+    },
+    {
+      "epoch": 57.00478632478632,
+      "grad_norm": 0.0063064745627343655,
+      "learning_rate": 4.7245963912630586e-06,
+      "loss": 0.0004,
+      "step": 13450
+    },
+    {
+      "epoch": 57.005213675213675,
+      "grad_norm": 0.001134122721850872,
+      "learning_rate": 4.719848053181387e-06,
+      "loss": 0.0,
+      "step": 13460
+    },
+    {
+      "epoch": 57.005641025641026,
+      "grad_norm": 0.01796836219727993,
+      "learning_rate": 4.715099715099716e-06,
+      "loss": 0.0001,
+      "step": 13470
+    },
+    {
+      "epoch": 57.00606837606838,
+      "grad_norm": 0.00409694854170084,
+      "learning_rate": 4.710351377018044e-06,
+      "loss": 0.0002,
+      "step": 13480
+    },
+    {
+      "epoch": 57.00649572649573,
+      "grad_norm": 0.00029308826196938753,
+      "learning_rate": 4.705603038936373e-06,
+      "loss": 0.0033,
+      "step": 13490
+    },
+    {
+      "epoch": 57.00692307692308,
+      "grad_norm": 0.0007244080770760775,
+      "learning_rate": 4.700854700854701e-06,
+      "loss": 0.0,
+      "step": 13500
+    },
+    {
+      "epoch": 57.007350427350424,
+      "grad_norm": 0.012751772068440914,
+      "learning_rate": 4.69610636277303e-06,
+      "loss": 0.0001,
+      "step": 13510
+    },
+    {
+      "epoch": 57.007777777777775,
+      "grad_norm": 0.0007130270823836327,
+      "learning_rate": 4.691358024691358e-06,
+      "loss": 0.0,
+      "step": 13520
+    },
+    {
+      "epoch": 57.00820512820513,
+      "grad_norm": 0.0003145153750665486,
+      "learning_rate": 4.6866096866096875e-06,
+      "loss": 0.1948,
+      "step": 13530
+    },
+    {
+      "epoch": 57.00863247863248,
+      "grad_norm": 0.00031738655525259674,
+      "learning_rate": 4.681861348528016e-06,
+      "loss": 0.0,
+      "step": 13540
+    },
+    {
+      "epoch": 57.00905982905983,
+      "grad_norm": 0.0002995093527715653,
+      "learning_rate": 4.677113010446344e-06,
+      "loss": 0.0001,
+      "step": 13550
+    },
+    {
+      "epoch": 57.00948717948718,
+      "grad_norm": 0.00032525527058169246,
+      "learning_rate": 4.672364672364672e-06,
+      "loss": 0.0,
+      "step": 13560
+    },
+    {
+      "epoch": 57.00991452991453,
+      "grad_norm": 0.00266217440366745,
+      "learning_rate": 4.6676163342830015e-06,
+      "loss": 0.0,
+      "step": 13570
+    },
+    {
+      "epoch": 57.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 5.7661638259887695,
+      "eval_runtime": 33.6667,
+      "eval_samples_per_second": 0.743,
+      "eval_steps_per_second": 0.743,
+      "step": 13572
+    },
+    {
+      "epoch": 58.00034188034188,
+      "grad_norm": 956.22998046875,
+      "learning_rate": 4.66286799620133e-06,
+      "loss": 0.9196,
+      "step": 13580
+    },
+    {
+      "epoch": 58.00076923076923,
+      "grad_norm": 0.0004830555117223412,
+      "learning_rate": 4.658119658119659e-06,
+      "loss": 0.0,
+      "step": 13590
+    },
+    {
+      "epoch": 58.00119658119658,
+      "grad_norm": 0.00030270041315816343,
+      "learning_rate": 4.653371320037987e-06,
+      "loss": 0.0,
+      "step": 13600
+    },
+    {
+      "epoch": 58.00162393162393,
+      "grad_norm": 0.0006561644258908927,
+      "learning_rate": 4.6486229819563155e-06,
+      "loss": 0.0,
+      "step": 13610
+    },
+    {
+      "epoch": 58.00205128205128,
+      "grad_norm": 0.0006711311871185899,
+      "learning_rate": 4.643874643874644e-06,
+      "loss": 0.0,
+      "step": 13620
+    },
+    {
+      "epoch": 58.002478632478635,
+      "grad_norm": 0.0025408940855413675,
+      "learning_rate": 4.639126305792973e-06,
+      "loss": 0.0,
+      "step": 13630
+    },
+    {
+      "epoch": 58.002905982905986,
+      "grad_norm": 0.0009754471248015761,
+      "learning_rate": 4.634377967711301e-06,
+      "loss": 0.0,
+      "step": 13640
+    },
+    {
+      "epoch": 58.00333333333333,
+      "grad_norm": 0.0024775350466370583,
+      "learning_rate": 4.62962962962963e-06,
+      "loss": 0.0,
+      "step": 13650
+    },
+    {
+      "epoch": 58.00376068376068,
+      "grad_norm": 0.007733345031738281,
+      "learning_rate": 4.624881291547959e-06,
+      "loss": 0.0001,
+      "step": 13660
+    },
+    {
+      "epoch": 58.00418803418803,
+      "grad_norm": 0.00038200884591788054,
+      "learning_rate": 4.620132953466287e-06,
+      "loss": 0.0,
+      "step": 13670
+    },
+    {
+      "epoch": 58.004615384615384,
+      "grad_norm": 0.00032710927189327776,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 0.0,
+      "step": 13680
+    },
+    {
+      "epoch": 58.005042735042736,
+      "grad_norm": 0.00046126110828481615,
+      "learning_rate": 4.6106362773029444e-06,
+      "loss": 0.0001,
+      "step": 13690
+    },
+    {
+      "epoch": 58.00547008547009,
+      "grad_norm": 0.0009468531352467835,
+      "learning_rate": 4.605887939221273e-06,
+      "loss": 0.0001,
+      "step": 13700
+    },
+    {
+      "epoch": 58.00589743589744,
+      "grad_norm": 0.0003751327458303422,
+      "learning_rate": 4.601139601139601e-06,
+      "loss": 0.0001,
+      "step": 13710
+    },
+    {
+      "epoch": 58.00632478632479,
+      "grad_norm": 0.0024412404745817184,
+      "learning_rate": 4.59639126305793e-06,
+      "loss": 0.0,
+      "step": 13720
+    },
+    {
+      "epoch": 58.006752136752134,
+      "grad_norm": 0.0011317358585074544,
+      "learning_rate": 4.5916429249762585e-06,
+      "loss": 0.0,
+      "step": 13730
+    },
+    {
+      "epoch": 58.007179487179485,
+      "grad_norm": 0.00213524978607893,
+      "learning_rate": 4.586894586894588e-06,
+      "loss": 0.0001,
+      "step": 13740
+    },
+    {
+      "epoch": 58.007606837606836,
+      "grad_norm": 0.00029954416095279157,
+      "learning_rate": 4.582146248812916e-06,
+      "loss": 0.0001,
+      "step": 13750
+    },
+    {
+      "epoch": 58.00803418803419,
+      "grad_norm": 0.004025555215775967,
+      "learning_rate": 4.577397910731244e-06,
+      "loss": 0.0,
+      "step": 13760
+    },
+    {
+      "epoch": 58.00846153846154,
+      "grad_norm": 0.0070886253379285336,
+      "learning_rate": 4.5726495726495725e-06,
+      "loss": 0.5328,
+      "step": 13770
+    },
+    {
+      "epoch": 58.00888888888889,
+      "grad_norm": 0.0006350260809995234,
+      "learning_rate": 4.567901234567902e-06,
+      "loss": 0.0,
+      "step": 13780
+    },
+    {
+      "epoch": 58.00931623931624,
+      "grad_norm": 89.9181900024414,
+      "learning_rate": 4.563152896486231e-06,
+      "loss": 0.0124,
+      "step": 13790
+    },
+    {
+      "epoch": 58.00974358974359,
+      "grad_norm": 0.000428661733167246,
+      "learning_rate": 4.558404558404559e-06,
+      "loss": 0.0001,
+      "step": 13800
+    },
+    {
+      "epoch": 58.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 5.446341037750244,
+      "eval_runtime": 33.6227,
+      "eval_samples_per_second": 0.744,
+      "eval_steps_per_second": 0.744,
+      "step": 13806
+    },
+    {
+      "epoch": 59.00017094017094,
+      "grad_norm": 0.0007079663337208331,
+      "learning_rate": 4.553656220322887e-06,
+      "loss": 0.0001,
+      "step": 13810
+    },
+    {
+      "epoch": 59.00059829059829,
+      "grad_norm": 0.0012396123493090272,
+      "learning_rate": 4.548907882241216e-06,
+      "loss": 1.3237,
+      "step": 13820
+    },
+    {
+      "epoch": 59.00102564102564,
+      "grad_norm": 0.005205500405281782,
+      "learning_rate": 4.544159544159544e-06,
+      "loss": 0.4377,
+      "step": 13830
+    },
+    {
+      "epoch": 59.00145299145299,
+      "grad_norm": 0.00035839254269376397,
+      "learning_rate": 4.539411206077873e-06,
+      "loss": 0.0083,
+      "step": 13840
+    },
+    {
+      "epoch": 59.001880341880344,
+      "grad_norm": 0.0005130122299306095,
+      "learning_rate": 4.5346628679962014e-06,
+      "loss": 0.0,
+      "step": 13850
+    },
+    {
+      "epoch": 59.002307692307696,
+      "grad_norm": 0.0014134275261312723,
+      "learning_rate": 4.5299145299145306e-06,
+      "loss": 0.0,
+      "step": 13860
+    },
+    {
+      "epoch": 59.00273504273504,
+      "grad_norm": 0.0625193640589714,
+      "learning_rate": 4.525166191832859e-06,
+      "loss": 0.0001,
+      "step": 13870
+    },
+    {
+      "epoch": 59.00316239316239,
+      "grad_norm": 0.0004801045579370111,
+      "learning_rate": 4.520417853751187e-06,
+      "loss": 0.0001,
+      "step": 13880
+    },
+    {
+      "epoch": 59.00358974358974,
+      "grad_norm": 0.0007496966863982379,
+      "learning_rate": 4.515669515669516e-06,
+      "loss": 0.1716,
+      "step": 13890
+    },
+    {
+      "epoch": 59.004017094017094,
+      "grad_norm": 0.0012893045786768198,
+      "learning_rate": 4.510921177587845e-06,
+      "loss": 0.0,
+      "step": 13900
+    },
+    {
+      "epoch": 59.004444444444445,
+      "grad_norm": 0.0003310160245746374,
+      "learning_rate": 4.506172839506173e-06,
+      "loss": 0.0001,
+      "step": 13910
+    },
+    {
+      "epoch": 59.0048717948718,
+      "grad_norm": 44.41610336303711,
+      "learning_rate": 4.501424501424502e-06,
+      "loss": 0.0076,
+      "step": 13920
+    },
+    {
+      "epoch": 59.00529914529915,
+      "grad_norm": 0.0027762632817029953,
+      "learning_rate": 4.49667616334283e-06,
+      "loss": 0.0,
+      "step": 13930
+    },
+    {
+      "epoch": 59.0057264957265,
+      "grad_norm": 0.0003408943011891097,
+      "learning_rate": 4.4919278252611595e-06,
+      "loss": 0.0,
+      "step": 13940
+    },
+    {
+      "epoch": 59.00615384615384,
+      "grad_norm": 0.0014558567199856043,
+      "learning_rate": 4.487179487179488e-06,
+      "loss": 0.0,
+      "step": 13950
+    },
+    {
+      "epoch": 59.006581196581195,
+      "grad_norm": 0.002356948796659708,
+      "learning_rate": 4.482431149097816e-06,
+      "loss": 0.0108,
+      "step": 13960
+    },
+    {
+      "epoch": 59.007008547008546,
+      "grad_norm": 0.0007559444638900459,
+      "learning_rate": 4.477682811016144e-06,
+      "loss": 0.6313,
+      "step": 13970
+    },
+    {
+      "epoch": 59.0074358974359,
+      "grad_norm": 0.004984420258551836,
+      "learning_rate": 4.4729344729344735e-06,
+      "loss": 0.1682,
+      "step": 13980
+    },
+    {
+      "epoch": 59.00786324786325,
+      "grad_norm": 0.0007952677551656961,
+      "learning_rate": 4.468186134852802e-06,
+      "loss": 0.9454,
+      "step": 13990
+    },
+    {
+      "epoch": 59.0082905982906,
+      "grad_norm": 0.02415173314511776,
+      "learning_rate": 4.463437796771131e-06,
+      "loss": 0.0001,
+      "step": 14000
+    },
+    {
+      "epoch": 59.00871794871795,
+      "grad_norm": 0.0003063694457523525,
+      "learning_rate": 4.458689458689459e-06,
+      "loss": 0.0001,
+      "step": 14010
+    },
+    {
+      "epoch": 59.0091452991453,
+      "grad_norm": 0.00032008966081775725,
+      "learning_rate": 4.4539411206077876e-06,
+      "loss": 0.4779,
+      "step": 14020
+    },
+    {
+      "epoch": 59.00957264957265,
+      "grad_norm": 0.00046775556984357536,
+      "learning_rate": 4.449192782526116e-06,
+      "loss": 0.0001,
+      "step": 14030
+    },
+    {
+      "epoch": 59.01,
+      "grad_norm": 0.06117213889956474,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.0021,
+      "step": 14040
+    },
+    {
+      "epoch": 59.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 5.957569599151611,
+      "eval_runtime": 36.0184,
+      "eval_samples_per_second": 0.694,
+      "eval_steps_per_second": 0.694,
+      "step": 14040
+    },
+    {
+      "epoch": 60.00042735042735,
+      "grad_norm": 0.0008756799506954849,
+      "learning_rate": 4.439696106362773e-06,
+      "loss": 0.6096,
+      "step": 14050
+    },
+    {
+      "epoch": 60.0008547008547,
+      "grad_norm": 945.39501953125,
+      "learning_rate": 4.4349477682811024e-06,
+      "loss": 0.5127,
+      "step": 14060
+    },
+    {
+      "epoch": 60.001282051282054,
+      "grad_norm": 0.0003207973495591432,
+      "learning_rate": 4.430199430199431e-06,
+      "loss": 0.0,
+      "step": 14070
+    },
+    {
+      "epoch": 60.001709401709405,
+      "grad_norm": 0.005959488917142153,
+      "learning_rate": 4.425451092117759e-06,
+      "loss": 0.0001,
+      "step": 14080
+    },
+    {
+      "epoch": 60.00213675213675,
+      "grad_norm": 0.0004705670871771872,
+      "learning_rate": 4.420702754036087e-06,
+      "loss": 0.0,
+      "step": 14090
+    },
+    {
+      "epoch": 60.0025641025641,
+      "grad_norm": 0.0004403094353619963,
+      "learning_rate": 4.4159544159544165e-06,
+      "loss": 0.0,
+      "step": 14100
+    },
+    {
+      "epoch": 60.00299145299145,
+      "grad_norm": 707.6570434570312,
+      "learning_rate": 4.411206077872745e-06,
+      "loss": 0.7011,
+      "step": 14110
+    },
+    {
+      "epoch": 60.0034188034188,
+      "grad_norm": 0.00034931182744912803,
+      "learning_rate": 4.406457739791074e-06,
+      "loss": 0.0011,
+      "step": 14120
+    },
+    {
+      "epoch": 60.003846153846155,
+      "grad_norm": 0.00047842093044891953,
+      "learning_rate": 4.401709401709402e-06,
+      "loss": 0.7669,
+      "step": 14130
+    },
+    {
+      "epoch": 60.004273504273506,
+      "grad_norm": 0.0005029200110584497,
+      "learning_rate": 4.3969610636277305e-06,
+      "loss": 0.6763,
+      "step": 14140
+    },
+    {
+      "epoch": 60.00470085470086,
+      "grad_norm": 0.0007298401324078441,
+      "learning_rate": 4.39221272554606e-06,
+      "loss": 0.0001,
+      "step": 14150
+    },
+    {
+      "epoch": 60.00512820512821,
+      "grad_norm": 0.08432787656784058,
+      "learning_rate": 4.387464387464388e-06,
+      "loss": 1.019,
+      "step": 14160
+    },
+    {
+      "epoch": 60.00555555555555,
+      "grad_norm": 0.011608411557972431,
+      "learning_rate": 4.382716049382716e-06,
+      "loss": 0.0,
+      "step": 14170
+    },
+    {
+      "epoch": 60.005982905982904,
+      "grad_norm": 0.0004716445109806955,
+      "learning_rate": 4.3779677113010445e-06,
+      "loss": 1.0345,
+      "step": 14180
+    },
+    {
+      "epoch": 60.006410256410255,
+      "grad_norm": 0.0010066544637084007,
+      "learning_rate": 4.373219373219374e-06,
+      "loss": 0.8799,
+      "step": 14190
+    },
+    {
+      "epoch": 60.00683760683761,
+      "grad_norm": 0.01125133316963911,
+      "learning_rate": 4.368471035137703e-06,
+      "loss": 0.0012,
+      "step": 14200
+    },
+    {
+      "epoch": 60.00726495726496,
+      "grad_norm": 0.0006305111455731094,
+      "learning_rate": 4.363722697056031e-06,
+      "loss": 0.7549,
+      "step": 14210
+    },
+    {
+      "epoch": 60.00769230769231,
+      "grad_norm": 0.003195669502019882,
+      "learning_rate": 4.358974358974359e-06,
+      "loss": 0.0,
+      "step": 14220
+    },
+    {
+      "epoch": 60.00811965811966,
+      "grad_norm": 0.20871169865131378,
+      "learning_rate": 4.354226020892688e-06,
+      "loss": 0.0753,
+      "step": 14230
+    },
+    {
+      "epoch": 60.00854700854701,
+      "grad_norm": 0.0661257803440094,
+      "learning_rate": 4.349477682811016e-06,
+      "loss": 0.0001,
+      "step": 14240
+    },
+    {
+      "epoch": 60.008974358974356,
+      "grad_norm": 0.0009889448992908,
+      "learning_rate": 4.344729344729345e-06,
+      "loss": 0.0002,
+      "step": 14250
+    },
+    {
+      "epoch": 60.00940170940171,
+      "grad_norm": 0.009586402215063572,
+      "learning_rate": 4.3399810066476735e-06,
+      "loss": 0.0004,
+      "step": 14260
+    },
+    {
+      "epoch": 60.00982905982906,
+      "grad_norm": 558.0621337890625,
+      "learning_rate": 4.335232668566003e-06,
+      "loss": 0.5042,
+      "step": 14270
+    },
+    {
+      "epoch": 60.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 5.9418511390686035,
+      "eval_runtime": 37.744,
+      "eval_samples_per_second": 0.662,
+      "eval_steps_per_second": 0.662,
+      "step": 14274
+    },
+    {
+      "epoch": 61.00025641025641,
+      "grad_norm": 0.00028759066481143236,
+      "learning_rate": 4.330484330484331e-06,
+      "loss": 0.9175,
+      "step": 14280
+    },
+    {
+      "epoch": 61.00068376068376,
+      "grad_norm": 0.007706713397055864,
+      "learning_rate": 4.325735992402659e-06,
+      "loss": 0.0001,
+      "step": 14290
+    },
+    {
+      "epoch": 61.00111111111111,
+      "grad_norm": 0.0005172354285605252,
+      "learning_rate": 4.3209876543209875e-06,
+      "loss": 0.0,
+      "step": 14300
+    },
+    {
+      "epoch": 61.00153846153846,
+      "grad_norm": 0.0005246769869700074,
+      "learning_rate": 4.316239316239317e-06,
+      "loss": 0.6762,
+      "step": 14310
+    },
+    {
+      "epoch": 61.00196581196581,
+      "grad_norm": 0.0011663725599646568,
+      "learning_rate": 4.311490978157645e-06,
+      "loss": 0.0,
+      "step": 14320
+    },
+    {
+      "epoch": 61.00239316239316,
+      "grad_norm": 0.005837870761752129,
+      "learning_rate": 4.306742640075974e-06,
+      "loss": 0.0907,
+      "step": 14330
+    },
+    {
+      "epoch": 61.00282051282051,
+      "grad_norm": 0.00042240836773999035,
+      "learning_rate": 4.301994301994302e-06,
+      "loss": 0.0,
+      "step": 14340
+    },
+    {
+      "epoch": 61.003247863247864,
+      "grad_norm": 0.0007428540848195553,
+      "learning_rate": 4.297245963912631e-06,
+      "loss": 0.0001,
+      "step": 14350
+    },
+    {
+      "epoch": 61.003675213675216,
+      "grad_norm": 0.0005527780158445239,
+      "learning_rate": 4.29249762583096e-06,
+      "loss": 0.0,
+      "step": 14360
+    },
+    {
+      "epoch": 61.00410256410257,
+      "grad_norm": 0.001011466607451439,
+      "learning_rate": 4.287749287749288e-06,
+      "loss": 0.0001,
+      "step": 14370
+    },
+    {
+      "epoch": 61.00452991452991,
+      "grad_norm": 17.799911499023438,
+      "learning_rate": 4.283000949667616e-06,
+      "loss": 0.0566,
+      "step": 14380
+    },
+    {
+      "epoch": 61.00495726495726,
+      "grad_norm": 20.882823944091797,
+      "learning_rate": 4.2782526115859455e-06,
+      "loss": 0.0043,
+      "step": 14390
+    },
+    {
+      "epoch": 61.005384615384614,
+      "grad_norm": 0.00030632095877081156,
+      "learning_rate": 4.273504273504274e-06,
+      "loss": 0.0004,
+      "step": 14400
+    },
+    {
+      "epoch": 61.005811965811965,
+      "grad_norm": 0.000464937009382993,
+      "learning_rate": 4.268755935422603e-06,
+      "loss": 0.0001,
+      "step": 14410
+    },
+    {
+      "epoch": 61.006239316239316,
+      "grad_norm": 0.0014673583209514618,
+      "learning_rate": 4.264007597340931e-06,
+      "loss": 0.0001,
+      "step": 14420
+    },
+    {
+      "epoch": 61.00666666666667,
+      "grad_norm": 0.0011050072498619556,
+      "learning_rate": 4.2592592592592596e-06,
+      "loss": 0.0,
+      "step": 14430
+    },
+    {
+      "epoch": 61.00709401709402,
+      "grad_norm": 408.4623107910156,
+      "learning_rate": 4.254510921177588e-06,
+      "loss": 0.5193,
+      "step": 14440
+    },
+    {
+      "epoch": 61.00752136752137,
+      "grad_norm": 0.0010529108112677932,
+      "learning_rate": 4.249762583095917e-06,
+      "loss": 0.0011,
+      "step": 14450
+    },
+    {
+      "epoch": 61.007948717948715,
+      "grad_norm": 0.0004740229342132807,
+      "learning_rate": 4.245014245014245e-06,
+      "loss": 0.0001,
+      "step": 14460
+    },
+    {
+      "epoch": 61.008376068376066,
+      "grad_norm": 0.0004031808057334274,
+      "learning_rate": 4.2402659069325745e-06,
+      "loss": 0.0,
+      "step": 14470
+    },
+    {
+      "epoch": 61.00880341880342,
+      "grad_norm": 0.00028951760032214224,
+      "learning_rate": 4.235517568850903e-06,
+      "loss": 0.0,
+      "step": 14480
+    },
+    {
+      "epoch": 61.00923076923077,
+      "grad_norm": 0.00033271079882979393,
+      "learning_rate": 4.230769230769231e-06,
+      "loss": 0.0001,
+      "step": 14490
+    },
+    {
+      "epoch": 61.00965811965812,
+      "grad_norm": 41.438385009765625,
+      "learning_rate": 4.226020892687559e-06,
+      "loss": 0.0053,
+      "step": 14500
+    },
+    {
+      "epoch": 61.01,
+      "eval_accuracy": 0.48,
+      "eval_loss": 5.297748565673828,
+      "eval_runtime": 37.9181,
+      "eval_samples_per_second": 0.659,
+      "eval_steps_per_second": 0.659,
+      "step": 14508
+    },
+    {
+      "epoch": 62.00008547008547,
+      "grad_norm": 0.0023157300893217325,
+      "learning_rate": 4.221272554605888e-06,
+      "loss": 0.0,
+      "step": 14510
+    },
+    {
+      "epoch": 62.00051282051282,
+      "grad_norm": 0.000285146088572219,
+      "learning_rate": 4.216524216524217e-06,
+      "loss": 0.0,
+      "step": 14520
+    },
+    {
+      "epoch": 62.00094017094017,
+      "grad_norm": 0.00043018904398195446,
+      "learning_rate": 4.211775878442546e-06,
+      "loss": 0.0,
+      "step": 14530
+    },
+    {
+      "epoch": 62.00136752136752,
+      "grad_norm": 0.0006764789577573538,
+      "learning_rate": 4.207027540360874e-06,
+      "loss": 0.2522,
+      "step": 14540
+    },
+    {
+      "epoch": 62.00179487179487,
+      "grad_norm": 0.0003662327362690121,
+      "learning_rate": 4.2022792022792025e-06,
+      "loss": 0.0,
+      "step": 14550
+    },
+    {
+      "epoch": 62.00222222222222,
+      "grad_norm": 0.0002613349643070251,
+      "learning_rate": 4.197530864197531e-06,
+      "loss": 0.0001,
+      "step": 14560
+    },
+    {
+      "epoch": 62.002649572649574,
+      "grad_norm": 0.0002579360152594745,
+      "learning_rate": 4.19278252611586e-06,
+      "loss": 0.0001,
+      "step": 14570
+    },
+    {
+      "epoch": 62.003076923076925,
+      "grad_norm": 0.002828446449711919,
+      "learning_rate": 4.188034188034188e-06,
+      "loss": 0.0001,
+      "step": 14580
+    },
+    {
+      "epoch": 62.00350427350428,
+      "grad_norm": 0.02188822254538536,
+      "learning_rate": 4.1832858499525166e-06,
+      "loss": 0.0,
+      "step": 14590
+    },
+    {
+      "epoch": 62.00393162393162,
+      "grad_norm": 0.0006723613478243351,
+      "learning_rate": 4.178537511870846e-06,
+      "loss": 0.0,
+      "step": 14600
+    },
+    {
+      "epoch": 62.00435897435897,
+      "grad_norm": 0.0004710882785730064,
+      "learning_rate": 4.173789173789174e-06,
+      "loss": 0.0,
+      "step": 14610
+    },
+    {
+      "epoch": 62.00478632478632,
+      "grad_norm": 0.0002582537126727402,
+      "learning_rate": 4.169040835707503e-06,
+      "loss": 0.0,
+      "step": 14620
+    },
+    {
+      "epoch": 62.005213675213675,
+      "grad_norm": 0.0019376453710719943,
+      "learning_rate": 4.1642924976258314e-06,
+      "loss": 0.0001,
+      "step": 14630
+    },
+    {
+      "epoch": 62.005641025641026,
+      "grad_norm": 0.00032024920801632106,
+      "learning_rate": 4.15954415954416e-06,
+      "loss": 0.0,
+      "step": 14640
+    },
+    {
+      "epoch": 62.00606837606838,
+      "grad_norm": 0.00024245944223366678,
+      "learning_rate": 4.154795821462488e-06,
+      "loss": 0.0,
+      "step": 14650
+    },
+    {
+      "epoch": 62.00649572649573,
+      "grad_norm": 0.0006904223700985312,
+      "learning_rate": 4.150047483380817e-06,
+      "loss": 0.7856,
+      "step": 14660
+    },
+    {
+      "epoch": 62.00692307692308,
+      "grad_norm": 0.00029405683744698763,
+      "learning_rate": 4.145299145299146e-06,
+      "loss": 0.0632,
+      "step": 14670
+    },
+    {
+      "epoch": 62.007350427350424,
+      "grad_norm": 0.0007452707504853606,
+      "learning_rate": 4.140550807217475e-06,
+      "loss": 0.0,
+      "step": 14680
+    },
+    {
+      "epoch": 62.007777777777775,
+      "grad_norm": 559.78662109375,
+      "learning_rate": 4.135802469135803e-06,
+      "loss": 0.3101,
+      "step": 14690
+    },
+    {
+      "epoch": 62.00820512820513,
+      "grad_norm": 0.02688130922615528,
+      "learning_rate": 4.131054131054131e-06,
+      "loss": 0.0001,
+      "step": 14700
+    },
+    {
+      "epoch": 62.00863247863248,
+      "grad_norm": 2.4815587997436523,
+      "learning_rate": 4.1263057929724595e-06,
+      "loss": 0.0005,
+      "step": 14710
+    },
+    {
+      "epoch": 62.00905982905983,
+      "grad_norm": 0.004581361077725887,
+      "learning_rate": 4.121557454890789e-06,
+      "loss": 0.0031,
+      "step": 14720
+    },
+    {
+      "epoch": 62.00948717948718,
+      "grad_norm": 0.017405280843377113,
+      "learning_rate": 4.116809116809117e-06,
+      "loss": 0.308,
+      "step": 14730
+    },
+    {
+      "epoch": 62.00991452991453,
+      "grad_norm": 0.001918576774187386,
+      "learning_rate": 4.112060778727446e-06,
+      "loss": 0.0,
+      "step": 14740
+    },
+    {
+      "epoch": 62.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 5.854069232940674,
+      "eval_runtime": 32.1126,
+      "eval_samples_per_second": 0.779,
+      "eval_steps_per_second": 0.779,
+      "step": 14742
+    },
+    {
+      "epoch": 63.00034188034188,
+      "grad_norm": 0.004541910719126463,
+      "learning_rate": 4.107312440645774e-06,
+      "loss": 0.0,
+      "step": 14750
+    },
+    {
+      "epoch": 63.00076923076923,
+      "grad_norm": 0.0010953666642308235,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 0.0001,
+      "step": 14760
+    },
+    {
+      "epoch": 63.00119658119658,
+      "grad_norm": 0.001070062629878521,
+      "learning_rate": 4.097815764482431e-06,
+      "loss": 0.0001,
+      "step": 14770
+    },
+    {
+      "epoch": 63.00162393162393,
+      "grad_norm": 0.001508062588982284,
+      "learning_rate": 4.09306742640076e-06,
+      "loss": 0.0001,
+      "step": 14780
+    },
+    {
+      "epoch": 63.00205128205128,
+      "grad_norm": 0.0003479434235487133,
+      "learning_rate": 4.088319088319088e-06,
+      "loss": 0.0,
+      "step": 14790
+    },
+    {
+      "epoch": 63.002478632478635,
+      "grad_norm": 0.0033716338220983744,
+      "learning_rate": 4.0835707502374176e-06,
+      "loss": 0.0001,
+      "step": 14800
+    },
+    {
+      "epoch": 63.002905982905986,
+      "grad_norm": 0.0008653479162603617,
+      "learning_rate": 4.078822412155746e-06,
+      "loss": 0.0173,
+      "step": 14810
+    },
+    {
+      "epoch": 63.00333333333333,
+      "grad_norm": 0.00030156198772601783,
+      "learning_rate": 4.074074074074074e-06,
+      "loss": 0.0,
+      "step": 14820
+    },
+    {
+      "epoch": 63.00376068376068,
+      "grad_norm": 0.0004595111822709441,
+      "learning_rate": 4.069325735992403e-06,
+      "loss": 0.0,
+      "step": 14830
+    },
+    {
+      "epoch": 63.00418803418803,
+      "grad_norm": 0.0010495609603822231,
+      "learning_rate": 4.064577397910732e-06,
+      "loss": 0.0,
+      "step": 14840
+    },
+    {
+      "epoch": 63.004615384615384,
+      "grad_norm": 0.004493600223213434,
+      "learning_rate": 4.05982905982906e-06,
+      "loss": 0.6305,
+      "step": 14850
+    },
+    {
+      "epoch": 63.005042735042736,
+      "grad_norm": 0.0505562499165535,
+      "learning_rate": 4.055080721747389e-06,
+      "loss": 0.0016,
+      "step": 14860
+    },
+    {
+      "epoch": 63.00547008547009,
+      "grad_norm": 0.002653477480635047,
+      "learning_rate": 4.050332383665717e-06,
+      "loss": 0.0001,
+      "step": 14870
+    },
+    {
+      "epoch": 63.00589743589744,
+      "grad_norm": 0.0020723838824778795,
+      "learning_rate": 4.0455840455840465e-06,
+      "loss": 0.0001,
+      "step": 14880
+    },
+    {
+      "epoch": 63.00632478632479,
+      "grad_norm": 0.0002628070651553571,
+      "learning_rate": 4.040835707502375e-06,
+      "loss": 0.0004,
+      "step": 14890
+    },
+    {
+      "epoch": 63.006752136752134,
+      "grad_norm": 0.0003271848545409739,
+      "learning_rate": 4.036087369420703e-06,
+      "loss": 0.0001,
+      "step": 14900
+    },
+    {
+      "epoch": 63.007179487179485,
+      "grad_norm": 0.0002233712439192459,
+      "learning_rate": 4.031339031339031e-06,
+      "loss": 0.0013,
+      "step": 14910
+    },
+    {
+      "epoch": 63.007606837606836,
+      "grad_norm": 0.000736577610950917,
+      "learning_rate": 4.02659069325736e-06,
+      "loss": 0.0,
+      "step": 14920
+    },
+    {
+      "epoch": 63.00803418803419,
+      "grad_norm": 0.0002595110272523016,
+      "learning_rate": 4.021842355175689e-06,
+      "loss": 0.0,
+      "step": 14930
+    },
+    {
+      "epoch": 63.00846153846154,
+      "grad_norm": 0.00025847507640719414,
+      "learning_rate": 4.017094017094018e-06,
+      "loss": 0.0,
+      "step": 14940
+    },
+    {
+      "epoch": 63.00888888888889,
+      "grad_norm": 0.0003213146119378507,
+      "learning_rate": 4.012345679012346e-06,
+      "loss": 0.0,
+      "step": 14950
+    },
+    {
+      "epoch": 63.00931623931624,
+      "grad_norm": 111.42301177978516,
+      "learning_rate": 4.0075973409306745e-06,
+      "loss": 0.0102,
+      "step": 14960
+    },
+    {
+      "epoch": 63.00974358974359,
+      "grad_norm": 705.5123291015625,
+      "learning_rate": 4.002849002849003e-06,
+      "loss": 0.1555,
+      "step": 14970
+    },
+    {
+      "epoch": 63.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.5367207527160645,
+      "eval_runtime": 31.9232,
+      "eval_samples_per_second": 0.783,
+      "eval_steps_per_second": 0.783,
+      "step": 14976
+    },
+    {
+      "epoch": 64.00017094017095,
+      "grad_norm": 0.42229506373405457,
+      "learning_rate": 3.998100664767331e-06,
+      "loss": 0.1863,
+      "step": 14980
+    },
+    {
+      "epoch": 64.00059829059829,
+      "grad_norm": 0.00032686052145436406,
+      "learning_rate": 3.99335232668566e-06,
+      "loss": 0.0,
+      "step": 14990
+    },
+    {
+      "epoch": 64.00102564102563,
+      "grad_norm": 0.0032908704597502947,
+      "learning_rate": 3.9886039886039894e-06,
+      "loss": 0.0002,
+      "step": 15000
+    },
+    {
+      "epoch": 64.001452991453,
+      "grad_norm": 3.955153703689575,
+      "learning_rate": 3.983855650522318e-06,
+      "loss": 0.0007,
+      "step": 15010
+    },
+    {
+      "epoch": 64.00188034188034,
+      "grad_norm": 0.0003446421178523451,
+      "learning_rate": 3.979107312440646e-06,
+      "loss": 1.0107,
+      "step": 15020
+    },
+    {
+      "epoch": 64.0023076923077,
+      "grad_norm": 0.002301581669598818,
+      "learning_rate": 3.974358974358974e-06,
+      "loss": 0.0,
+      "step": 15030
+    },
+    {
+      "epoch": 64.00273504273504,
+      "grad_norm": 0.0007779121515341103,
+      "learning_rate": 3.9696106362773035e-06,
+      "loss": 0.7845,
+      "step": 15040
+    },
+    {
+      "epoch": 64.0031623931624,
+      "grad_norm": 0.11843976378440857,
+      "learning_rate": 3.964862298195632e-06,
+      "loss": 0.1222,
+      "step": 15050
+    },
+    {
+      "epoch": 64.00358974358974,
+      "grad_norm": 0.020959218963980675,
+      "learning_rate": 3.96011396011396e-06,
+      "loss": 0.0,
+      "step": 15060
+    },
+    {
+      "epoch": 64.0040170940171,
+      "grad_norm": 506.3919982910156,
+      "learning_rate": 3.955365622032289e-06,
+      "loss": 0.0546,
+      "step": 15070
+    },
+    {
+      "epoch": 64.00444444444445,
+      "grad_norm": 0.0002198971196776256,
+      "learning_rate": 3.9506172839506175e-06,
+      "loss": 0.0,
+      "step": 15080
+    },
+    {
+      "epoch": 64.00487179487179,
+      "grad_norm": 0.008080464787781239,
+      "learning_rate": 3.945868945868947e-06,
+      "loss": 0.0001,
+      "step": 15090
+    },
+    {
+      "epoch": 64.00529914529915,
+      "grad_norm": 0.002219392219558358,
+      "learning_rate": 3.941120607787275e-06,
+      "loss": 0.0007,
+      "step": 15100
+    },
+    {
+      "epoch": 64.00572649572649,
+      "grad_norm": 0.0003887212951667607,
+      "learning_rate": 3.936372269705603e-06,
+      "loss": 0.0016,
+      "step": 15110
+    },
+    {
+      "epoch": 64.00615384615385,
+      "grad_norm": 0.001071627251803875,
+      "learning_rate": 3.9316239316239315e-06,
+      "loss": 0.0,
+      "step": 15120
+    },
+    {
+      "epoch": 64.0065811965812,
+      "grad_norm": 0.00028002672479487956,
+      "learning_rate": 3.926875593542261e-06,
+      "loss": 0.0,
+      "step": 15130
+    },
+    {
+      "epoch": 64.00700854700855,
+      "grad_norm": 0.00031417288118973374,
+      "learning_rate": 3.922127255460589e-06,
+      "loss": 0.0,
+      "step": 15140
+    },
+    {
+      "epoch": 64.0074358974359,
+      "grad_norm": 0.0010667052119970322,
+      "learning_rate": 3.917378917378918e-06,
+      "loss": 0.0273,
+      "step": 15150
+    },
+    {
+      "epoch": 64.00786324786324,
+      "grad_norm": 0.0006762260454706848,
+      "learning_rate": 3.912630579297246e-06,
+      "loss": 0.051,
+      "step": 15160
+    },
+    {
+      "epoch": 64.0082905982906,
+      "grad_norm": 0.006420073565095663,
+      "learning_rate": 3.907882241215575e-06,
+      "loss": 0.0,
+      "step": 15170
+    },
+    {
+      "epoch": 64.00871794871794,
+      "grad_norm": 0.0005878534284420311,
+      "learning_rate": 3.903133903133903e-06,
+      "loss": 0.0,
+      "step": 15180
+    },
+    {
+      "epoch": 64.0091452991453,
+      "grad_norm": 0.0019342320738360286,
+      "learning_rate": 3.898385565052232e-06,
+      "loss": 0.0,
+      "step": 15190
+    },
+    {
+      "epoch": 64.00957264957265,
+      "grad_norm": 0.00036040143459104,
+      "learning_rate": 3.8936372269705604e-06,
+      "loss": 0.0002,
+      "step": 15200
+    },
+    {
+      "epoch": 64.01,
+      "grad_norm": 0.0003253432805649936,
+      "learning_rate": 3.88888888888889e-06,
+      "loss": 0.0081,
+      "step": 15210
+    },
+    {
+      "epoch": 64.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 5.480795383453369,
+      "eval_runtime": 31.9103,
+      "eval_samples_per_second": 0.783,
+      "eval_steps_per_second": 0.783,
+      "step": 15210
+    },
+    {
+      "epoch": 65.00042735042734,
+      "grad_norm": 0.0006786137819290161,
+      "learning_rate": 3.884140550807218e-06,
+      "loss": 0.0001,
+      "step": 15220
+    },
+    {
+      "epoch": 65.0008547008547,
+      "grad_norm": 0.00022978027118369937,
+      "learning_rate": 3.879392212725546e-06,
+      "loss": 0.0,
+      "step": 15230
+    },
+    {
+      "epoch": 65.00128205128205,
+      "grad_norm": 0.0007715200772508979,
+      "learning_rate": 3.8746438746438745e-06,
+      "loss": 0.003,
+      "step": 15240
+    },
+    {
+      "epoch": 65.0017094017094,
+      "grad_norm": 0.02402709610760212,
+      "learning_rate": 3.869895536562204e-06,
+      "loss": 0.0,
+      "step": 15250
+    },
+    {
+      "epoch": 65.00213675213675,
+      "grad_norm": 0.0015532065881416202,
+      "learning_rate": 3.865147198480532e-06,
+      "loss": 0.0197,
+      "step": 15260
+    },
+    {
+      "epoch": 65.00256410256411,
+      "grad_norm": 85.6712417602539,
+      "learning_rate": 3.860398860398861e-06,
+      "loss": 1.1,
+      "step": 15270
+    },
+    {
+      "epoch": 65.00299145299145,
+      "grad_norm": 0.0003184400557074696,
+      "learning_rate": 3.855650522317189e-06,
+      "loss": 0.0012,
+      "step": 15280
+    },
+    {
+      "epoch": 65.00341880341881,
+      "grad_norm": 0.0002915811201091856,
+      "learning_rate": 3.850902184235518e-06,
+      "loss": 0.0002,
+      "step": 15290
+    },
+    {
+      "epoch": 65.00384615384615,
+      "grad_norm": 0.001770331640727818,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 0.0023,
+      "step": 15300
+    },
+    {
+      "epoch": 65.0042735042735,
+      "grad_norm": 0.0037825191393494606,
+      "learning_rate": 3.841405508072175e-06,
+      "loss": 0.0,
+      "step": 15310
+    },
+    {
+      "epoch": 65.00470085470086,
+      "grad_norm": 0.00039001365075819194,
+      "learning_rate": 3.836657169990503e-06,
+      "loss": 0.0,
+      "step": 15320
+    },
+    {
+      "epoch": 65.0051282051282,
+      "grad_norm": 0.029118506237864494,
+      "learning_rate": 3.8319088319088325e-06,
+      "loss": 0.0001,
+      "step": 15330
+    },
+    {
+      "epoch": 65.00555555555556,
+      "grad_norm": 0.004345818888396025,
+      "learning_rate": 3.827160493827161e-06,
+      "loss": 0.0001,
+      "step": 15340
+    },
+    {
+      "epoch": 65.0059829059829,
+      "grad_norm": 0.0013317528646439314,
+      "learning_rate": 3.82241215574549e-06,
+      "loss": 0.0,
+      "step": 15350
+    },
+    {
+      "epoch": 65.00641025641026,
+      "grad_norm": 0.00033316414919681847,
+      "learning_rate": 3.817663817663818e-06,
+      "loss": 0.0,
+      "step": 15360
+    },
+    {
+      "epoch": 65.0068376068376,
+      "grad_norm": 0.006611653603613377,
+      "learning_rate": 3.8129154795821466e-06,
+      "loss": 0.0001,
+      "step": 15370
+    },
+    {
+      "epoch": 65.00726495726495,
+      "grad_norm": 0.0014399340143427253,
+      "learning_rate": 3.808167141500475e-06,
+      "loss": 0.0024,
+      "step": 15380
+    },
+    {
+      "epoch": 65.00769230769231,
+      "grad_norm": 0.000310930801788345,
+      "learning_rate": 3.8034188034188036e-06,
+      "loss": 0.0001,
+      "step": 15390
+    },
+    {
+      "epoch": 65.00811965811965,
+      "grad_norm": 0.0017023945692926645,
+      "learning_rate": 3.7986704653371327e-06,
+      "loss": 0.0,
+      "step": 15400
+    },
+    {
+      "epoch": 65.00854700854701,
+      "grad_norm": 0.01913696527481079,
+      "learning_rate": 3.793922127255461e-06,
+      "loss": 0.0,
+      "step": 15410
+    },
+    {
+      "epoch": 65.00897435897436,
+      "grad_norm": 0.00039296565228141844,
+      "learning_rate": 3.7891737891737893e-06,
+      "loss": 0.0,
+      "step": 15420
+    },
+    {
+      "epoch": 65.00940170940171,
+      "grad_norm": 0.00039896511589176953,
+      "learning_rate": 3.784425451092118e-06,
+      "loss": 1.0074,
+      "step": 15430
+    },
+    {
+      "epoch": 65.00982905982906,
+      "grad_norm": 0.0005804836982861161,
+      "learning_rate": 3.7796771130104463e-06,
+      "loss": 0.0008,
+      "step": 15440
+    },
+    {
+      "epoch": 65.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 5.881788730621338,
+      "eval_runtime": 35.8586,
+      "eval_samples_per_second": 0.697,
+      "eval_steps_per_second": 0.697,
+      "step": 15444
+    },
+    {
+      "epoch": 66.00025641025641,
+      "grad_norm": 0.0012229053536430001,
+      "learning_rate": 3.774928774928775e-06,
+      "loss": 0.0001,
+      "step": 15450
+    },
+    {
+      "epoch": 66.00068376068376,
+      "grad_norm": 0.0003784662112593651,
+      "learning_rate": 3.770180436847104e-06,
+      "loss": 0.0001,
+      "step": 15460
+    },
+    {
+      "epoch": 66.00111111111111,
+      "grad_norm": 0.0005255657597444952,
+      "learning_rate": 3.7654320987654325e-06,
+      "loss": 0.0001,
+      "step": 15470
+    },
+    {
+      "epoch": 66.00153846153846,
+      "grad_norm": 0.00034700758988037705,
+      "learning_rate": 3.760683760683761e-06,
+      "loss": 0.0005,
+      "step": 15480
+    },
+    {
+      "epoch": 66.00196581196582,
+      "grad_norm": 0.00021708589338231832,
+      "learning_rate": 3.7559354226020895e-06,
+      "loss": 0.0,
+      "step": 15490
+    },
+    {
+      "epoch": 66.00239316239316,
+      "grad_norm": 0.0038388799875974655,
+      "learning_rate": 3.751187084520418e-06,
+      "loss": 0.0,
+      "step": 15500
+    },
+    {
+      "epoch": 66.0028205128205,
+      "grad_norm": 0.0007136081694625318,
+      "learning_rate": 3.746438746438747e-06,
+      "loss": 0.0001,
+      "step": 15510
+    },
+    {
+      "epoch": 66.00324786324786,
+      "grad_norm": 0.0002888034505303949,
+      "learning_rate": 3.7416904083570752e-06,
+      "loss": 0.0,
+      "step": 15520
+    },
+    {
+      "epoch": 66.00367521367521,
+      "grad_norm": 720.2564697265625,
+      "learning_rate": 3.736942070275404e-06,
+      "loss": 0.2916,
+      "step": 15530
+    },
+    {
+      "epoch": 66.00410256410257,
+      "grad_norm": 0.00027234895969741046,
+      "learning_rate": 3.7321937321937323e-06,
+      "loss": 0.0003,
+      "step": 15540
+    },
+    {
+      "epoch": 66.00452991452991,
+      "grad_norm": 0.0012739448575302958,
+      "learning_rate": 3.727445394112061e-06,
+      "loss": 0.0,
+      "step": 15550
+    },
+    {
+      "epoch": 66.00495726495727,
+      "grad_norm": 0.0005203865002840757,
+      "learning_rate": 3.7226970560303897e-06,
+      "loss": 0.0,
+      "step": 15560
+    },
+    {
+      "epoch": 66.00538461538461,
+      "grad_norm": 0.0003277095383964479,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 0.0,
+      "step": 15570
+    },
+    {
+      "epoch": 66.00581196581197,
+      "grad_norm": 0.00044921765220351517,
+      "learning_rate": 3.7132003798670467e-06,
+      "loss": 0.0,
+      "step": 15580
+    },
+    {
+      "epoch": 66.00623931623932,
+      "grad_norm": 0.0003073754196520895,
+      "learning_rate": 3.7084520417853754e-06,
+      "loss": 0.0,
+      "step": 15590
+    },
+    {
+      "epoch": 66.00666666666666,
+      "grad_norm": 0.00030189528479240835,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 0.0001,
+      "step": 15600
+    },
+    {
+      "epoch": 66.00709401709402,
+      "grad_norm": 0.0005707856616936624,
+      "learning_rate": 3.698955365622033e-06,
+      "loss": 0.0001,
+      "step": 15610
+    },
+    {
+      "epoch": 66.00752136752136,
+      "grad_norm": 0.00040740257827565074,
+      "learning_rate": 3.694207027540361e-06,
+      "loss": 0.0,
+      "step": 15620
+    },
+    {
+      "epoch": 66.00794871794872,
+      "grad_norm": 0.00039323364035226405,
+      "learning_rate": 3.68945868945869e-06,
+      "loss": 0.0,
+      "step": 15630
+    },
+    {
+      "epoch": 66.00837606837607,
+      "grad_norm": 0.0014208744978532195,
+      "learning_rate": 3.684710351377018e-06,
+      "loss": 0.0,
+      "step": 15640
+    },
+    {
+      "epoch": 66.00880341880342,
+      "grad_norm": 0.0007801980245858431,
+      "learning_rate": 3.679962013295347e-06,
+      "loss": 0.0001,
+      "step": 15650
+    },
+    {
+      "epoch": 66.00923076923077,
+      "grad_norm": 0.0017435887129977345,
+      "learning_rate": 3.6752136752136756e-06,
+      "loss": 0.0001,
+      "step": 15660
+    },
+    {
+      "epoch": 66.00965811965811,
+      "grad_norm": 0.03651350364089012,
+      "learning_rate": 3.6704653371320044e-06,
+      "loss": 0.0,
+      "step": 15670
+    },
+    {
+      "epoch": 66.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.4378485679626465,
+      "eval_runtime": 33.2393,
+      "eval_samples_per_second": 0.752,
+      "eval_steps_per_second": 0.752,
+      "step": 15678
+    },
+    {
+      "epoch": 67.00008547008547,
+      "grad_norm": 0.00035544627462513745,
+      "learning_rate": 3.6657169990503327e-06,
+      "loss": 0.0001,
+      "step": 15680
+    },
+    {
+      "epoch": 67.00051282051282,
+      "grad_norm": 0.00033005906152538955,
+      "learning_rate": 3.6609686609686614e-06,
+      "loss": 0.0,
+      "step": 15690
+    },
+    {
+      "epoch": 67.00094017094017,
+      "grad_norm": 0.0003103645867668092,
+      "learning_rate": 3.6562203228869897e-06,
+      "loss": 0.0,
+      "step": 15700
+    },
+    {
+      "epoch": 67.00136752136753,
+      "grad_norm": 0.0008121723076328635,
+      "learning_rate": 3.651471984805318e-06,
+      "loss": 0.0,
+      "step": 15710
+    },
+    {
+      "epoch": 67.00179487179487,
+      "grad_norm": 0.0004981961101293564,
+      "learning_rate": 3.646723646723647e-06,
+      "loss": 0.0,
+      "step": 15720
+    },
+    {
+      "epoch": 67.00222222222222,
+      "grad_norm": 0.0003290316089987755,
+      "learning_rate": 3.641975308641976e-06,
+      "loss": 0.0,
+      "step": 15730
+    },
+    {
+      "epoch": 67.00264957264957,
+      "grad_norm": 0.0010973671451210976,
+      "learning_rate": 3.637226970560304e-06,
+      "loss": 0.0,
+      "step": 15740
+    },
+    {
+      "epoch": 67.00307692307692,
+      "grad_norm": 0.0006377913523465395,
+      "learning_rate": 3.632478632478633e-06,
+      "loss": 0.0,
+      "step": 15750
+    },
+    {
+      "epoch": 67.00350427350428,
+      "grad_norm": 0.0005305053782649338,
+      "learning_rate": 3.627730294396961e-06,
+      "loss": 0.0,
+      "step": 15760
+    },
+    {
+      "epoch": 67.00393162393162,
+      "grad_norm": 0.0019237673841416836,
+      "learning_rate": 3.6229819563152903e-06,
+      "loss": 0.0006,
+      "step": 15770
+    },
+    {
+      "epoch": 67.00435897435898,
+      "grad_norm": 0.0016178624937310815,
+      "learning_rate": 3.6182336182336186e-06,
+      "loss": 0.0,
+      "step": 15780
+    },
+    {
+      "epoch": 67.00478632478632,
+      "grad_norm": 0.00046172275324352086,
+      "learning_rate": 3.6134852801519473e-06,
+      "loss": 0.0,
+      "step": 15790
+    },
+    {
+      "epoch": 67.00521367521368,
+      "grad_norm": 0.0002967897162307054,
+      "learning_rate": 3.6087369420702756e-06,
+      "loss": 0.0001,
+      "step": 15800
+    },
+    {
+      "epoch": 67.00564102564103,
+      "grad_norm": 0.0002238056476926431,
+      "learning_rate": 3.603988603988604e-06,
+      "loss": 0.0941,
+      "step": 15810
+    },
+    {
+      "epoch": 67.00606837606837,
+      "grad_norm": 0.00390687957406044,
+      "learning_rate": 3.599240265906933e-06,
+      "loss": 0.001,
+      "step": 15820
+    },
+    {
+      "epoch": 67.00649572649573,
+      "grad_norm": 0.0019423745106905699,
+      "learning_rate": 3.5944919278252618e-06,
+      "loss": 0.0001,
+      "step": 15830
+    },
+    {
+      "epoch": 67.00692307692307,
+      "grad_norm": 0.00027149979723617435,
+      "learning_rate": 3.58974358974359e-06,
+      "loss": 0.0,
+      "step": 15840
+    },
+    {
+      "epoch": 67.00735042735043,
+      "grad_norm": 0.0007080771028995514,
+      "learning_rate": 3.5849952516619184e-06,
+      "loss": 0.0,
+      "step": 15850
+    },
+    {
+      "epoch": 67.00777777777778,
+      "grad_norm": 0.0003103041963186115,
+      "learning_rate": 3.580246913580247e-06,
+      "loss": 0.0135,
+      "step": 15860
+    },
+    {
+      "epoch": 67.00820512820513,
+      "grad_norm": 0.001910716644488275,
+      "learning_rate": 3.5754985754985762e-06,
+      "loss": 0.0,
+      "step": 15870
+    },
+    {
+      "epoch": 67.00863247863248,
+      "grad_norm": 904.8418579101562,
+      "learning_rate": 3.5707502374169045e-06,
+      "loss": 0.3092,
+      "step": 15880
+    },
+    {
+      "epoch": 67.00905982905982,
+      "grad_norm": 0.0006746945437043905,
+      "learning_rate": 3.566001899335233e-06,
+      "loss": 0.0,
+      "step": 15890
+    },
+    {
+      "epoch": 67.00948717948718,
+      "grad_norm": 0.0018293018219992518,
+      "learning_rate": 3.5612535612535615e-06,
+      "loss": 0.0001,
+      "step": 15900
+    },
+    {
+      "epoch": 67.00991452991452,
+      "grad_norm": 0.007653082720935345,
+      "learning_rate": 3.55650522317189e-06,
+      "loss": 0.0,
+      "step": 15910
+    },
+    {
+      "epoch": 67.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 5.659733772277832,
+      "eval_runtime": 35.0596,
+      "eval_samples_per_second": 0.713,
+      "eval_steps_per_second": 0.713,
+      "step": 15912
+    },
+    {
+      "epoch": 68.00034188034188,
+      "grad_norm": 0.0002990284119732678,
+      "learning_rate": 3.551756885090219e-06,
+      "loss": 0.0272,
+      "step": 15920
+    },
+    {
+      "epoch": 68.00076923076924,
+      "grad_norm": 0.0034541944041848183,
+      "learning_rate": 3.5470085470085473e-06,
+      "loss": 0.0001,
+      "step": 15930
+    },
+    {
+      "epoch": 68.00119658119658,
+      "grad_norm": 0.00031465632491745055,
+      "learning_rate": 3.542260208926876e-06,
+      "loss": 0.0,
+      "step": 15940
+    },
+    {
+      "epoch": 68.00162393162393,
+      "grad_norm": 0.0003612114815041423,
+      "learning_rate": 3.5375118708452043e-06,
+      "loss": 0.0002,
+      "step": 15950
+    },
+    {
+      "epoch": 68.00205128205128,
+      "grad_norm": 0.0013253976358100772,
+      "learning_rate": 3.532763532763533e-06,
+      "loss": 0.0,
+      "step": 15960
+    },
+    {
+      "epoch": 68.00247863247863,
+      "grad_norm": 0.0014574190136045218,
+      "learning_rate": 3.5280151946818613e-06,
+      "loss": 0.0,
+      "step": 15970
+    },
+    {
+      "epoch": 68.00290598290599,
+      "grad_norm": 0.0002389967121416703,
+      "learning_rate": 3.5232668566001904e-06,
+      "loss": 0.0003,
+      "step": 15980
+    },
+    {
+      "epoch": 68.00333333333333,
+      "grad_norm": 0.0006442684680223465,
+      "learning_rate": 3.5185185185185187e-06,
+      "loss": 0.0,
+      "step": 15990
+    },
+    {
+      "epoch": 68.00376068376069,
+      "grad_norm": 0.0004441600467544049,
+      "learning_rate": 3.5137701804368475e-06,
+      "loss": 0.0,
+      "step": 16000
+    },
+    {
+      "epoch": 68.00418803418803,
+      "grad_norm": 0.0006568527896888554,
+      "learning_rate": 3.5090218423551758e-06,
+      "loss": 0.0,
+      "step": 16010
+    },
+    {
+      "epoch": 68.00461538461539,
+      "grad_norm": 0.0003113803395535797,
+      "learning_rate": 3.5042735042735045e-06,
+      "loss": 0.0,
+      "step": 16020
+    },
+    {
+      "epoch": 68.00504273504274,
+      "grad_norm": 0.0012534753186628222,
+      "learning_rate": 3.499525166191833e-06,
+      "loss": 0.0,
+      "step": 16030
+    },
+    {
+      "epoch": 68.00547008547008,
+      "grad_norm": 0.0026429889257997274,
+      "learning_rate": 3.494776828110162e-06,
+      "loss": 0.2954,
+      "step": 16040
+    },
+    {
+      "epoch": 68.00589743589744,
+      "grad_norm": 0.0012878195848315954,
+      "learning_rate": 3.4900284900284902e-06,
+      "loss": 0.0,
+      "step": 16050
+    },
+    {
+      "epoch": 68.00632478632478,
+      "grad_norm": 0.0004001953057013452,
+      "learning_rate": 3.485280151946819e-06,
+      "loss": 0.4867,
+      "step": 16060
+    },
+    {
+      "epoch": 68.00675213675214,
+      "grad_norm": 0.004468402359634638,
+      "learning_rate": 3.4805318138651472e-06,
+      "loss": 0.0006,
+      "step": 16070
+    },
+    {
+      "epoch": 68.00717948717949,
+      "grad_norm": 0.002943785861134529,
+      "learning_rate": 3.4757834757834764e-06,
+      "loss": 0.0002,
+      "step": 16080
+    },
+    {
+      "epoch": 68.00760683760684,
+      "grad_norm": 0.002336192177608609,
+      "learning_rate": 3.4710351377018047e-06,
+      "loss": 0.8271,
+      "step": 16090
+    },
+    {
+      "epoch": 68.00803418803419,
+      "grad_norm": 0.0003368109464645386,
+      "learning_rate": 3.4662867996201334e-06,
+      "loss": 0.3306,
+      "step": 16100
+    },
+    {
+      "epoch": 68.00846153846153,
+      "grad_norm": 0.005345764569938183,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 0.498,
+      "step": 16110
+    },
+    {
+      "epoch": 68.00888888888889,
+      "grad_norm": 0.005093185696750879,
+      "learning_rate": 3.4567901234567904e-06,
+      "loss": 0.0001,
+      "step": 16120
+    },
+    {
+      "epoch": 68.00931623931623,
+      "grad_norm": 0.00032298368751071393,
+      "learning_rate": 3.452041785375119e-06,
+      "loss": 0.0,
+      "step": 16130
+    },
+    {
+      "epoch": 68.0097435897436,
+      "grad_norm": 0.0008845495176501572,
+      "learning_rate": 3.447293447293448e-06,
+      "loss": 0.0,
+      "step": 16140
+    },
+    {
+      "epoch": 68.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 5.819715976715088,
+      "eval_runtime": 32.8708,
+      "eval_samples_per_second": 0.761,
+      "eval_steps_per_second": 0.761,
+      "step": 16146
+    },
+    {
+      "epoch": 69.00017094017095,
+      "grad_norm": 0.003377682762220502,
+      "learning_rate": 3.442545109211776e-06,
+      "loss": 0.0002,
+      "step": 16150
+    },
+    {
+      "epoch": 69.00059829059829,
+      "grad_norm": 0.0002674728457350284,
+      "learning_rate": 3.437796771130105e-06,
+      "loss": 0.0,
+      "step": 16160
+    },
+    {
+      "epoch": 69.00102564102563,
+      "grad_norm": 0.0016994690522551537,
+      "learning_rate": 3.433048433048433e-06,
+      "loss": 0.0,
+      "step": 16170
+    },
+    {
+      "epoch": 69.001452991453,
+      "grad_norm": 0.00035284223849885166,
+      "learning_rate": 3.4283000949667615e-06,
+      "loss": 0.0,
+      "step": 16180
+    },
+    {
+      "epoch": 69.00188034188034,
+      "grad_norm": 0.00023986499581951648,
+      "learning_rate": 3.4235517568850906e-06,
+      "loss": 0.0,
+      "step": 16190
+    },
+    {
+      "epoch": 69.0023076923077,
+      "grad_norm": 0.0023895238991826773,
+      "learning_rate": 3.4188034188034193e-06,
+      "loss": 0.0,
+      "step": 16200
+    },
+    {
+      "epoch": 69.00273504273504,
+      "grad_norm": 0.0018031138461083174,
+      "learning_rate": 3.4140550807217476e-06,
+      "loss": 0.0,
+      "step": 16210
+    },
+    {
+      "epoch": 69.0031623931624,
+      "grad_norm": 0.003573950147256255,
+      "learning_rate": 3.409306742640076e-06,
+      "loss": 0.0,
+      "step": 16220
+    },
+    {
+      "epoch": 69.00358974358974,
+      "grad_norm": 0.0015754502965137362,
+      "learning_rate": 3.4045584045584046e-06,
+      "loss": 0.0,
+      "step": 16230
+    },
+    {
+      "epoch": 69.0040170940171,
+      "grad_norm": 0.0012102372711524367,
+      "learning_rate": 3.3998100664767338e-06,
+      "loss": 0.0,
+      "step": 16240
+    },
+    {
+      "epoch": 69.00444444444445,
+      "grad_norm": 0.0002841241657733917,
+      "learning_rate": 3.395061728395062e-06,
+      "loss": 0.0,
+      "step": 16250
+    },
+    {
+      "epoch": 69.00487179487179,
+      "grad_norm": 0.01132782083004713,
+      "learning_rate": 3.3903133903133904e-06,
+      "loss": 0.0001,
+      "step": 16260
+    },
+    {
+      "epoch": 69.00529914529915,
+      "grad_norm": 0.00031476642470806837,
+      "learning_rate": 3.385565052231719e-06,
+      "loss": 0.0001,
+      "step": 16270
+    },
+    {
+      "epoch": 69.00572649572649,
+      "grad_norm": 0.000270717719104141,
+      "learning_rate": 3.3808167141500474e-06,
+      "loss": 0.0,
+      "step": 16280
+    },
+    {
+      "epoch": 69.00615384615385,
+      "grad_norm": 0.00036470277700573206,
+      "learning_rate": 3.3760683760683765e-06,
+      "loss": 0.0,
+      "step": 16290
+    },
+    {
+      "epoch": 69.0065811965812,
+      "grad_norm": 0.001150319236330688,
+      "learning_rate": 3.371320037986705e-06,
+      "loss": 0.0,
+      "step": 16300
+    },
+    {
+      "epoch": 69.00700854700855,
+      "grad_norm": 0.0003526509099174291,
+      "learning_rate": 3.3665716999050336e-06,
+      "loss": 0.0,
+      "step": 16310
+    },
+    {
+      "epoch": 69.0074358974359,
+      "grad_norm": 0.0005036306101828814,
+      "learning_rate": 3.361823361823362e-06,
+      "loss": 0.0,
+      "step": 16320
+    },
+    {
+      "epoch": 69.00786324786324,
+      "grad_norm": 0.0007515393663197756,
+      "learning_rate": 3.3570750237416906e-06,
+      "loss": 0.0,
+      "step": 16330
+    },
+    {
+      "epoch": 69.0082905982906,
+      "grad_norm": 0.0001931802835315466,
+      "learning_rate": 3.3523266856600197e-06,
+      "loss": 0.0,
+      "step": 16340
+    },
+    {
+      "epoch": 69.00871794871794,
+      "grad_norm": 0.00042826347635127604,
+      "learning_rate": 3.347578347578348e-06,
+      "loss": 0.0001,
+      "step": 16350
+    },
+    {
+      "epoch": 69.0091452991453,
+      "grad_norm": 0.00022397881548386067,
+      "learning_rate": 3.3428300094966763e-06,
+      "loss": 0.0,
+      "step": 16360
+    },
+    {
+      "epoch": 69.00957264957265,
+      "grad_norm": 0.00021716070477850735,
+      "learning_rate": 3.338081671415005e-06,
+      "loss": 0.0,
+      "step": 16370
+    },
+    {
+      "epoch": 69.01,
+      "grad_norm": 0.00088693160796538,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.0061,
+      "step": 16380
+    },
+    {
+      "epoch": 69.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.014081001281738,
+      "eval_runtime": 33.8632,
+      "eval_samples_per_second": 0.738,
+      "eval_steps_per_second": 0.738,
+      "step": 16380
+    },
+    {
+      "epoch": 70.00042735042734,
+      "grad_norm": 0.0004485013196244836,
+      "learning_rate": 3.3285849952516625e-06,
+      "loss": 0.0,
+      "step": 16390
+    },
+    {
+      "epoch": 70.0008547008547,
+      "grad_norm": 0.0004966802662238479,
+      "learning_rate": 3.3238366571699908e-06,
+      "loss": 0.0,
+      "step": 16400
+    },
+    {
+      "epoch": 70.00128205128205,
+      "grad_norm": 0.0006574277649633586,
+      "learning_rate": 3.3190883190883195e-06,
+      "loss": 0.001,
+      "step": 16410
+    },
+    {
+      "epoch": 70.0017094017094,
+      "grad_norm": 0.21187789738178253,
+      "learning_rate": 3.3143399810066478e-06,
+      "loss": 0.0001,
+      "step": 16420
+    },
+    {
+      "epoch": 70.00213675213675,
+      "grad_norm": 0.00020025823323521763,
+      "learning_rate": 3.3095916429249765e-06,
+      "loss": 0.0,
+      "step": 16430
+    },
+    {
+      "epoch": 70.00256410256411,
+      "grad_norm": 0.0003043335455004126,
+      "learning_rate": 3.304843304843305e-06,
+      "loss": 0.0,
+      "step": 16440
+    },
+    {
+      "epoch": 70.00299145299145,
+      "grad_norm": 0.0003146652306895703,
+      "learning_rate": 3.300094966761634e-06,
+      "loss": 0.0,
+      "step": 16450
+    },
+    {
+      "epoch": 70.00341880341881,
+      "grad_norm": 0.00840036105364561,
+      "learning_rate": 3.2953466286799622e-06,
+      "loss": 1.0593,
+      "step": 16460
+    },
+    {
+      "epoch": 70.00384615384615,
+      "grad_norm": 0.0003192836884409189,
+      "learning_rate": 3.290598290598291e-06,
+      "loss": 0.0,
+      "step": 16470
+    },
+    {
+      "epoch": 70.0042735042735,
+      "grad_norm": 0.0012308260193094611,
+      "learning_rate": 3.2858499525166193e-06,
+      "loss": 0.0,
+      "step": 16480
+    },
+    {
+      "epoch": 70.00470085470086,
+      "grad_norm": 0.0008993004448711872,
+      "learning_rate": 3.281101614434948e-06,
+      "loss": 0.0,
+      "step": 16490
+    },
+    {
+      "epoch": 70.0051282051282,
+      "grad_norm": 0.0010926368413493037,
+      "learning_rate": 3.2763532763532767e-06,
+      "loss": 0.0,
+      "step": 16500
+    },
+    {
+      "epoch": 70.00555555555556,
+      "grad_norm": 0.0005557582480832934,
+      "learning_rate": 3.2716049382716054e-06,
+      "loss": 0.0001,
+      "step": 16510
+    },
+    {
+      "epoch": 70.0059829059829,
+      "grad_norm": 0.0008928438182920218,
+      "learning_rate": 3.2668566001899337e-06,
+      "loss": 0.0,
+      "step": 16520
+    },
+    {
+      "epoch": 70.00641025641026,
+      "grad_norm": 0.0003105592622887343,
+      "learning_rate": 3.2621082621082624e-06,
+      "loss": 0.0001,
+      "step": 16530
+    },
+    {
+      "epoch": 70.0068376068376,
+      "grad_norm": 0.0003956545260734856,
+      "learning_rate": 3.2573599240265907e-06,
+      "loss": 0.0,
+      "step": 16540
+    },
+    {
+      "epoch": 70.00726495726495,
+      "grad_norm": 0.0002564324822742492,
+      "learning_rate": 3.25261158594492e-06,
+      "loss": 0.0066,
+      "step": 16550
+    },
+    {
+      "epoch": 70.00769230769231,
+      "grad_norm": 0.00025351883959956467,
+      "learning_rate": 3.247863247863248e-06,
+      "loss": 0.0,
+      "step": 16560
+    },
+    {
+      "epoch": 70.00811965811965,
+      "grad_norm": 0.00046608541742898524,
+      "learning_rate": 3.243114909781577e-06,
+      "loss": 0.0023,
+      "step": 16570
+    },
+    {
+      "epoch": 70.00854700854701,
+      "grad_norm": 0.002306720009073615,
+      "learning_rate": 3.238366571699905e-06,
+      "loss": 0.0,
+      "step": 16580
+    },
+    {
+      "epoch": 70.00897435897436,
+      "grad_norm": 0.0005604177713394165,
+      "learning_rate": 3.2336182336182335e-06,
+      "loss": 0.0,
+      "step": 16590
+    },
+    {
+      "epoch": 70.00940170940171,
+      "grad_norm": 0.0008777379989624023,
+      "learning_rate": 3.2288698955365626e-06,
+      "loss": 0.0,
+      "step": 16600
+    },
+    {
+      "epoch": 70.00982905982906,
+      "grad_norm": 0.0015941828023642302,
+      "learning_rate": 3.2241215574548913e-06,
+      "loss": 0.0001,
+      "step": 16610
+    },
+    {
+      "epoch": 70.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.244938373565674,
+      "eval_runtime": 34.6306,
+      "eval_samples_per_second": 0.722,
+      "eval_steps_per_second": 0.722,
+      "step": 16614
+    },
+    {
+      "epoch": 71.00025641025641,
+      "grad_norm": 0.0021248271223157644,
+      "learning_rate": 3.2193732193732196e-06,
+      "loss": 0.7973,
+      "step": 16620
+    },
+    {
+      "epoch": 71.00068376068376,
+      "grad_norm": 0.00018868227198254317,
+      "learning_rate": 3.214624881291548e-06,
+      "loss": 0.0,
+      "step": 16630
+    },
+    {
+      "epoch": 71.00111111111111,
+      "grad_norm": 921.1409301757812,
+      "learning_rate": 3.2098765432098767e-06,
+      "loss": 0.3031,
+      "step": 16640
+    },
+    {
+      "epoch": 71.00153846153846,
+      "grad_norm": 0.0012714448384940624,
+      "learning_rate": 3.205128205128206e-06,
+      "loss": 0.0,
+      "step": 16650
+    },
+    {
+      "epoch": 71.00196581196582,
+      "grad_norm": 0.0014924101997166872,
+      "learning_rate": 3.200379867046534e-06,
+      "loss": 0.0079,
+      "step": 16660
+    },
+    {
+      "epoch": 71.00239316239316,
+      "grad_norm": 0.00040770156192593277,
+      "learning_rate": 3.195631528964863e-06,
+      "loss": 0.0018,
+      "step": 16670
+    },
+    {
+      "epoch": 71.0028205128205,
+      "grad_norm": 0.00022908284154254943,
+      "learning_rate": 3.190883190883191e-06,
+      "loss": 0.002,
+      "step": 16680
+    },
+    {
+      "epoch": 71.00324786324786,
+      "grad_norm": 0.0016254654619842768,
+      "learning_rate": 3.1861348528015194e-06,
+      "loss": 0.7552,
+      "step": 16690
+    },
+    {
+      "epoch": 71.00367521367521,
+      "grad_norm": 0.0006611282587982714,
+      "learning_rate": 3.181386514719848e-06,
+      "loss": 0.0972,
+      "step": 16700
+    },
+    {
+      "epoch": 71.00410256410257,
+      "grad_norm": 2.668363332748413,
+      "learning_rate": 3.1766381766381773e-06,
+      "loss": 0.0006,
+      "step": 16710
+    },
+    {
+      "epoch": 71.00452991452991,
+      "grad_norm": 0.0007531665614806116,
+      "learning_rate": 3.1718898385565056e-06,
+      "loss": 0.0,
+      "step": 16720
+    },
+    {
+      "epoch": 71.00495726495727,
+      "grad_norm": 0.0007412447594106197,
+      "learning_rate": 3.167141500474834e-06,
+      "loss": 0.0,
+      "step": 16730
+    },
+    {
+      "epoch": 71.00538461538461,
+      "grad_norm": 0.0008477046503685415,
+      "learning_rate": 3.1623931623931626e-06,
+      "loss": 0.9046,
+      "step": 16740
+    },
+    {
+      "epoch": 71.00581196581197,
+      "grad_norm": 146.11962890625,
+      "learning_rate": 3.157644824311491e-06,
+      "loss": 0.3816,
+      "step": 16750
+    },
+    {
+      "epoch": 71.00623931623932,
+      "grad_norm": 0.0002443613775540143,
+      "learning_rate": 3.15289648622982e-06,
+      "loss": 0.2428,
+      "step": 16760
+    },
+    {
+      "epoch": 71.00666666666666,
+      "grad_norm": 0.00033410079777240753,
+      "learning_rate": 3.1481481481481483e-06,
+      "loss": 0.0,
+      "step": 16770
+    },
+    {
+      "epoch": 71.00709401709402,
+      "grad_norm": 0.002586389658972621,
+      "learning_rate": 3.143399810066477e-06,
+      "loss": 0.8823,
+      "step": 16780
+    },
+    {
+      "epoch": 71.00752136752136,
+      "grad_norm": 0.001075676642358303,
+      "learning_rate": 3.1386514719848053e-06,
+      "loss": 0.0,
+      "step": 16790
+    },
+    {
+      "epoch": 71.00794871794872,
+      "grad_norm": 0.002573596313595772,
+      "learning_rate": 3.133903133903134e-06,
+      "loss": 0.0,
+      "step": 16800
+    },
+    {
+      "epoch": 71.00837606837607,
+      "grad_norm": 0.00022963988885749131,
+      "learning_rate": 3.1291547958214628e-06,
+      "loss": 0.0,
+      "step": 16810
+    },
+    {
+      "epoch": 71.00880341880342,
+      "grad_norm": 0.0005840660887770355,
+      "learning_rate": 3.1244064577397915e-06,
+      "loss": 0.0,
+      "step": 16820
+    },
+    {
+      "epoch": 71.00923076923077,
+      "grad_norm": 0.00029322231421247125,
+      "learning_rate": 3.11965811965812e-06,
+      "loss": 0.0,
+      "step": 16830
+    },
+    {
+      "epoch": 71.00965811965811,
+      "grad_norm": 0.00025132184964604676,
+      "learning_rate": 3.1149097815764485e-06,
+      "loss": 0.0001,
+      "step": 16840
+    },
+    {
+      "epoch": 71.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.252962589263916,
+      "eval_runtime": 33.4983,
+      "eval_samples_per_second": 0.746,
+      "eval_steps_per_second": 0.746,
+      "step": 16848
+    },
+    {
+      "epoch": 72.00008547008547,
+      "grad_norm": 0.0019571471493691206,
+      "learning_rate": 3.110161443494777e-06,
+      "loss": 0.0,
+      "step": 16850
+    },
+    {
+      "epoch": 72.00051282051282,
+      "grad_norm": 0.0005339854396879673,
+      "learning_rate": 3.105413105413106e-06,
+      "loss": 0.0001,
+      "step": 16860
+    },
+    {
+      "epoch": 72.00094017094017,
+      "grad_norm": 0.00019883255299646407,
+      "learning_rate": 3.1006647673314343e-06,
+      "loss": 0.0,
+      "step": 16870
+    },
+    {
+      "epoch": 72.00136752136753,
+      "grad_norm": 0.00024648249382153153,
+      "learning_rate": 3.095916429249763e-06,
+      "loss": 0.0003,
+      "step": 16880
+    },
+    {
+      "epoch": 72.00179487179487,
+      "grad_norm": 0.0015446147881448269,
+      "learning_rate": 3.0911680911680913e-06,
+      "loss": 0.0006,
+      "step": 16890
+    },
+    {
+      "epoch": 72.00222222222222,
+      "grad_norm": 0.0021758642978966236,
+      "learning_rate": 3.08641975308642e-06,
+      "loss": 0.0001,
+      "step": 16900
+    },
+    {
+      "epoch": 72.00264957264957,
+      "grad_norm": 0.0003914159897249192,
+      "learning_rate": 3.0816714150047487e-06,
+      "loss": 0.0,
+      "step": 16910
+    },
+    {
+      "epoch": 72.00307692307692,
+      "grad_norm": 0.0009236117475666106,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 0.0,
+      "step": 16920
+    },
+    {
+      "epoch": 72.00350427350428,
+      "grad_norm": 0.0002460191317368299,
+      "learning_rate": 3.0721747388414057e-06,
+      "loss": 0.0,
+      "step": 16930
+    },
+    {
+      "epoch": 72.00393162393162,
+      "grad_norm": 0.00033975299447774887,
+      "learning_rate": 3.0674264007597345e-06,
+      "loss": 0.0,
+      "step": 16940
+    },
+    {
+      "epoch": 72.00435897435898,
+      "grad_norm": 0.001101949717849493,
+      "learning_rate": 3.0626780626780627e-06,
+      "loss": 0.0,
+      "step": 16950
+    },
+    {
+      "epoch": 72.00478632478632,
+      "grad_norm": 0.0031472702976316214,
+      "learning_rate": 3.0579297245963915e-06,
+      "loss": 0.5291,
+      "step": 16960
+    },
+    {
+      "epoch": 72.00521367521368,
+      "grad_norm": 0.00030517615959979594,
+      "learning_rate": 3.05318138651472e-06,
+      "loss": 0.0,
+      "step": 16970
+    },
+    {
+      "epoch": 72.00564102564103,
+      "grad_norm": 0.0015221843495965004,
+      "learning_rate": 3.048433048433049e-06,
+      "loss": 0.0004,
+      "step": 16980
+    },
+    {
+      "epoch": 72.00606837606837,
+      "grad_norm": 0.0034548670519143343,
+      "learning_rate": 3.043684710351377e-06,
+      "loss": 0.0,
+      "step": 16990
+    },
+    {
+      "epoch": 72.00649572649573,
+      "grad_norm": 0.0006523687625303864,
+      "learning_rate": 3.038936372269706e-06,
+      "loss": 0.9208,
+      "step": 17000
+    },
+    {
+      "epoch": 72.00692307692307,
+      "grad_norm": 0.0015028053894639015,
+      "learning_rate": 3.0341880341880342e-06,
+      "loss": 0.0,
+      "step": 17010
+    },
+    {
+      "epoch": 72.00735042735043,
+      "grad_norm": 0.00025994013412855566,
+      "learning_rate": 3.0294396961063634e-06,
+      "loss": 0.0,
+      "step": 17020
+    },
+    {
+      "epoch": 72.00777777777778,
+      "grad_norm": 0.000636533775832504,
+      "learning_rate": 3.0246913580246917e-06,
+      "loss": 0.0303,
+      "step": 17030
+    },
+    {
+      "epoch": 72.00820512820513,
+      "grad_norm": 0.003830546513199806,
+      "learning_rate": 3.0199430199430204e-06,
+      "loss": 0.0001,
+      "step": 17040
+    },
+    {
+      "epoch": 72.00863247863248,
+      "grad_norm": 0.00017824105452746153,
+      "learning_rate": 3.0151946818613487e-06,
+      "loss": 0.0,
+      "step": 17050
+    },
+    {
+      "epoch": 72.00905982905982,
+      "grad_norm": 0.00022492320567835122,
+      "learning_rate": 3.010446343779677e-06,
+      "loss": 0.0,
+      "step": 17060
+    },
+    {
+      "epoch": 72.00948717948718,
+      "grad_norm": 0.0020875674672424793,
+      "learning_rate": 3.005698005698006e-06,
+      "loss": 0.0,
+      "step": 17070
+    },
+    {
+      "epoch": 72.00991452991452,
+      "grad_norm": 0.0006881671724840999,
+      "learning_rate": 3.000949667616335e-06,
+      "loss": 0.0,
+      "step": 17080
+    },
+    {
+      "epoch": 72.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 5.765503406524658,
+      "eval_runtime": 33.3353,
+      "eval_samples_per_second": 0.75,
+      "eval_steps_per_second": 0.75,
+      "step": 17082
+    },
+    {
+      "epoch": 73.00034188034188,
+      "grad_norm": 0.0012791932094842196,
+      "learning_rate": 2.996201329534663e-06,
+      "loss": 0.0,
+      "step": 17090
+    },
+    {
+      "epoch": 73.00076923076924,
+      "grad_norm": 0.00039805949199944735,
+      "learning_rate": 2.9914529914529914e-06,
+      "loss": 0.0,
+      "step": 17100
+    },
+    {
+      "epoch": 73.00119658119658,
+      "grad_norm": 0.0010755633702501655,
+      "learning_rate": 2.98670465337132e-06,
+      "loss": 0.0001,
+      "step": 17110
+    },
+    {
+      "epoch": 73.00162393162393,
+      "grad_norm": 0.07058065384626389,
+      "learning_rate": 2.9819563152896493e-06,
+      "loss": 0.0,
+      "step": 17120
+    },
+    {
+      "epoch": 73.00205128205128,
+      "grad_norm": 0.00041162193519994617,
+      "learning_rate": 2.9772079772079776e-06,
+      "loss": 0.0,
+      "step": 17130
+    },
+    {
+      "epoch": 73.00247863247863,
+      "grad_norm": 0.0027573970146477222,
+      "learning_rate": 2.972459639126306e-06,
+      "loss": 0.0,
+      "step": 17140
+    },
+    {
+      "epoch": 73.00290598290599,
+      "grad_norm": 0.0003702895774040371,
+      "learning_rate": 2.9677113010446346e-06,
+      "loss": 0.0,
+      "step": 17150
+    },
+    {
+      "epoch": 73.00333333333333,
+      "grad_norm": 0.0007125766715034842,
+      "learning_rate": 2.962962962962963e-06,
+      "loss": 0.0,
+      "step": 17160
+    },
+    {
+      "epoch": 73.00376068376069,
+      "grad_norm": 0.00024414356448687613,
+      "learning_rate": 2.9582146248812916e-06,
+      "loss": 0.0,
+      "step": 17170
+    },
+    {
+      "epoch": 73.00418803418803,
+      "grad_norm": 0.0009582286584191024,
+      "learning_rate": 2.9534662867996203e-06,
+      "loss": 0.0001,
+      "step": 17180
+    },
+    {
+      "epoch": 73.00461538461539,
+      "grad_norm": 0.00019269179028924555,
+      "learning_rate": 2.948717948717949e-06,
+      "loss": 0.0,
+      "step": 17190
+    },
+    {
+      "epoch": 73.00504273504274,
+      "grad_norm": 0.0013847124064341187,
+      "learning_rate": 2.9439696106362774e-06,
+      "loss": 0.0,
+      "step": 17200
+    },
+    {
+      "epoch": 73.00547008547008,
+      "grad_norm": 0.8285155892372131,
+      "learning_rate": 2.939221272554606e-06,
+      "loss": 0.0002,
+      "step": 17210
+    },
+    {
+      "epoch": 73.00589743589744,
+      "grad_norm": 0.0001845570222940296,
+      "learning_rate": 2.9344729344729344e-06,
+      "loss": 0.0,
+      "step": 17220
+    },
+    {
+      "epoch": 73.00632478632478,
+      "grad_norm": 561.99072265625,
+      "learning_rate": 2.9297245963912635e-06,
+      "loss": 0.6593,
+      "step": 17230
+    },
+    {
+      "epoch": 73.00675213675214,
+      "grad_norm": 0.17753881216049194,
+      "learning_rate": 2.924976258309592e-06,
+      "loss": 0.0001,
+      "step": 17240
+    },
+    {
+      "epoch": 73.00717948717949,
+      "grad_norm": 0.00024875238887034357,
+      "learning_rate": 2.9202279202279205e-06,
+      "loss": 0.0,
+      "step": 17250
+    },
+    {
+      "epoch": 73.00760683760684,
+      "grad_norm": 0.016402428969740868,
+      "learning_rate": 2.915479582146249e-06,
+      "loss": 0.0039,
+      "step": 17260
+    },
+    {
+      "epoch": 73.00803418803419,
+      "grad_norm": 0.001006082515232265,
+      "learning_rate": 2.9107312440645776e-06,
+      "loss": 0.0038,
+      "step": 17270
+    },
+    {
+      "epoch": 73.00846153846153,
+      "grad_norm": 0.0013266198802739382,
+      "learning_rate": 2.9059829059829063e-06,
+      "loss": 0.0113,
+      "step": 17280
+    },
+    {
+      "epoch": 73.00888888888889,
+      "grad_norm": 0.00038516995846293867,
+      "learning_rate": 2.901234567901235e-06,
+      "loss": 0.0,
+      "step": 17290
+    },
+    {
+      "epoch": 73.00931623931623,
+      "grad_norm": 0.0010296168038621545,
+      "learning_rate": 2.8964862298195633e-06,
+      "loss": 0.0,
+      "step": 17300
+    },
+    {
+      "epoch": 73.0097435897436,
+      "grad_norm": 0.0006145525840111077,
+      "learning_rate": 2.891737891737892e-06,
+      "loss": 0.0,
+      "step": 17310
+    },
+    {
+      "epoch": 73.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.152116775512695,
+      "eval_runtime": 33.4645,
+      "eval_samples_per_second": 0.747,
+      "eval_steps_per_second": 0.747,
+      "step": 17316
+    },
+    {
+      "epoch": 74.00017094017095,
+      "grad_norm": 0.00017263834888581187,
+      "learning_rate": 2.8869895536562203e-06,
+      "loss": 0.0,
+      "step": 17320
+    },
+    {
+      "epoch": 74.00059829059829,
+      "grad_norm": 0.00031588933779858053,
+      "learning_rate": 2.8822412155745495e-06,
+      "loss": 0.0034,
+      "step": 17330
+    },
+    {
+      "epoch": 74.00102564102563,
+      "grad_norm": 0.0006647444679401815,
+      "learning_rate": 2.8774928774928778e-06,
+      "loss": 0.0,
+      "step": 17340
+    },
+    {
+      "epoch": 74.001452991453,
+      "grad_norm": 0.0013857238227501512,
+      "learning_rate": 2.8727445394112065e-06,
+      "loss": 0.0004,
+      "step": 17350
+    },
+    {
+      "epoch": 74.00188034188034,
+      "grad_norm": 0.003892462467774749,
+      "learning_rate": 2.8679962013295348e-06,
+      "loss": 0.0002,
+      "step": 17360
+    },
+    {
+      "epoch": 74.0023076923077,
+      "grad_norm": 0.0004585685092024505,
+      "learning_rate": 2.8632478632478635e-06,
+      "loss": 0.0,
+      "step": 17370
+    },
+    {
+      "epoch": 74.00273504273504,
+      "grad_norm": 0.00025991059374064207,
+      "learning_rate": 2.858499525166192e-06,
+      "loss": 0.0008,
+      "step": 17380
+    },
+    {
+      "epoch": 74.0031623931624,
+      "grad_norm": 0.0004791081009898335,
+      "learning_rate": 2.853751187084521e-06,
+      "loss": 0.0354,
+      "step": 17390
+    },
+    {
+      "epoch": 74.00358974358974,
+      "grad_norm": 0.0005468902527354658,
+      "learning_rate": 2.8490028490028492e-06,
+      "loss": 0.0001,
+      "step": 17400
+    },
+    {
+      "epoch": 74.0040170940171,
+      "grad_norm": 0.01083172857761383,
+      "learning_rate": 2.844254510921178e-06,
+      "loss": 0.0,
+      "step": 17410
+    },
+    {
+      "epoch": 74.00444444444445,
+      "grad_norm": 0.0011809701099991798,
+      "learning_rate": 2.8395061728395062e-06,
+      "loss": 0.0,
+      "step": 17420
+    },
+    {
+      "epoch": 74.00487179487179,
+      "grad_norm": 0.0013473471626639366,
+      "learning_rate": 2.8347578347578345e-06,
+      "loss": 0.0,
+      "step": 17430
+    },
+    {
+      "epoch": 74.00529914529915,
+      "grad_norm": 0.00039711128920316696,
+      "learning_rate": 2.8300094966761637e-06,
+      "loss": 0.0,
+      "step": 17440
+    },
+    {
+      "epoch": 74.00572649572649,
+      "grad_norm": 0.00017134596419055015,
+      "learning_rate": 2.8252611585944924e-06,
+      "loss": 0.0,
+      "step": 17450
+    },
+    {
+      "epoch": 74.00615384615385,
+      "grad_norm": 0.00042395462514832616,
+      "learning_rate": 2.8205128205128207e-06,
+      "loss": 0.0,
+      "step": 17460
+    },
+    {
+      "epoch": 74.0065811965812,
+      "grad_norm": 0.00016071122081484646,
+      "learning_rate": 2.815764482431149e-06,
+      "loss": 0.0,
+      "step": 17470
+    },
+    {
+      "epoch": 74.00700854700855,
+      "grad_norm": 0.20301441848278046,
+      "learning_rate": 2.8110161443494777e-06,
+      "loss": 0.0001,
+      "step": 17480
+    },
+    {
+      "epoch": 74.0074358974359,
+      "grad_norm": 0.0002695897710509598,
+      "learning_rate": 2.806267806267807e-06,
+      "loss": 0.0,
+      "step": 17490
+    },
+    {
+      "epoch": 74.00786324786324,
+      "grad_norm": 0.005475103389471769,
+      "learning_rate": 2.801519468186135e-06,
+      "loss": 0.0,
+      "step": 17500
+    },
+    {
+      "epoch": 74.0082905982906,
+      "grad_norm": 0.01020097080618143,
+      "learning_rate": 2.7967711301044635e-06,
+      "loss": 0.0,
+      "step": 17510
+    },
+    {
+      "epoch": 74.00871794871794,
+      "grad_norm": 0.00026164480368606746,
+      "learning_rate": 2.792022792022792e-06,
+      "loss": 0.0,
+      "step": 17520
+    },
+    {
+      "epoch": 74.0091452991453,
+      "grad_norm": 0.0005212483229115605,
+      "learning_rate": 2.7872744539411205e-06,
+      "loss": 0.0001,
+      "step": 17530
+    },
+    {
+      "epoch": 74.00957264957265,
+      "grad_norm": 0.00018099919543601573,
+      "learning_rate": 2.7825261158594496e-06,
+      "loss": 0.0,
+      "step": 17540
+    },
+    {
+      "epoch": 74.01,
+      "grad_norm": 0.00015604341751895845,
+      "learning_rate": 2.7777777777777783e-06,
+      "loss": 0.0,
+      "step": 17550
+    },
+    {
+      "epoch": 74.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 6.159724235534668,
+      "eval_runtime": 33.1573,
+      "eval_samples_per_second": 0.754,
+      "eval_steps_per_second": 0.754,
+      "step": 17550
+    },
+    {
+      "epoch": 75.00042735042734,
+      "grad_norm": 0.00023313055862672627,
+      "learning_rate": 2.7730294396961066e-06,
+      "loss": 0.0,
+      "step": 17560
+    },
+    {
+      "epoch": 75.0008547008547,
+      "grad_norm": 0.002859926549717784,
+      "learning_rate": 2.768281101614435e-06,
+      "loss": 0.0,
+      "step": 17570
+    },
+    {
+      "epoch": 75.00128205128205,
+      "grad_norm": 0.002907276153564453,
+      "learning_rate": 2.7635327635327636e-06,
+      "loss": 0.0001,
+      "step": 17580
+    },
+    {
+      "epoch": 75.0017094017094,
+      "grad_norm": 0.0013060672208666801,
+      "learning_rate": 2.758784425451093e-06,
+      "loss": 0.0,
+      "step": 17590
+    },
+    {
+      "epoch": 75.00213675213675,
+      "grad_norm": 0.0003227469860576093,
+      "learning_rate": 2.754036087369421e-06,
+      "loss": 0.0,
+      "step": 17600
+    },
+    {
+      "epoch": 75.00256410256411,
+      "grad_norm": 0.0003190785355400294,
+      "learning_rate": 2.7492877492877494e-06,
+      "loss": 0.0,
+      "step": 17610
+    },
+    {
+      "epoch": 75.00299145299145,
+      "grad_norm": 0.00020427798153832555,
+      "learning_rate": 2.744539411206078e-06,
+      "loss": 0.0,
+      "step": 17620
+    },
+    {
+      "epoch": 75.00341880341881,
+      "grad_norm": 0.053999051451683044,
+      "learning_rate": 2.7397910731244064e-06,
+      "loss": 0.0001,
+      "step": 17630
+    },
+    {
+      "epoch": 75.00384615384615,
+      "grad_norm": 0.0002565534086897969,
+      "learning_rate": 2.7350427350427355e-06,
+      "loss": 0.0,
+      "step": 17640
+    },
+    {
+      "epoch": 75.0042735042735,
+      "grad_norm": 0.00017350871348753572,
+      "learning_rate": 2.730294396961064e-06,
+      "loss": 0.0,
+      "step": 17650
+    },
+    {
+      "epoch": 75.00470085470086,
+      "grad_norm": 0.00023242604220286012,
+      "learning_rate": 2.7255460588793926e-06,
+      "loss": 0.0,
+      "step": 17660
+    },
+    {
+      "epoch": 75.0051282051282,
+      "grad_norm": 0.00020732257689815015,
+      "learning_rate": 2.720797720797721e-06,
+      "loss": 0.0,
+      "step": 17670
+    },
+    {
+      "epoch": 75.00555555555556,
+      "grad_norm": 0.00024015163944568485,
+      "learning_rate": 2.7160493827160496e-06,
+      "loss": 0.0,
+      "step": 17680
+    },
+    {
+      "epoch": 75.0059829059829,
+      "grad_norm": 0.00017238871078006923,
+      "learning_rate": 2.711301044634378e-06,
+      "loss": 0.0,
+      "step": 17690
+    },
+    {
+      "epoch": 75.00641025641026,
+      "grad_norm": 0.0001814868883229792,
+      "learning_rate": 2.706552706552707e-06,
+      "loss": 0.0002,
+      "step": 17700
+    },
+    {
+      "epoch": 75.0068376068376,
+      "grad_norm": 0.00020034310000482947,
+      "learning_rate": 2.7018043684710353e-06,
+      "loss": 0.0,
+      "step": 17710
+    },
+    {
+      "epoch": 75.00726495726495,
+      "grad_norm": 0.004398710560053587,
+      "learning_rate": 2.697056030389364e-06,
+      "loss": 0.0,
+      "step": 17720
+    },
+    {
+      "epoch": 75.00769230769231,
+      "grad_norm": 0.00018758252554107457,
+      "learning_rate": 2.6923076923076923e-06,
+      "loss": 0.0,
+      "step": 17730
+    },
+    {
+      "epoch": 75.00811965811965,
+      "grad_norm": 0.0014624909963458776,
+      "learning_rate": 2.687559354226021e-06,
+      "loss": 0.0,
+      "step": 17740
+    },
+    {
+      "epoch": 75.00854700854701,
+      "grad_norm": 0.0002650118840392679,
+      "learning_rate": 2.6828110161443498e-06,
+      "loss": 0.0,
+      "step": 17750
+    },
+    {
+      "epoch": 75.00897435897436,
+      "grad_norm": 0.004331326577812433,
+      "learning_rate": 2.6780626780626785e-06,
+      "loss": 0.0,
+      "step": 17760
+    },
+    {
+      "epoch": 75.00940170940171,
+      "grad_norm": 0.0013256591046229005,
+      "learning_rate": 2.673314339981007e-06,
+      "loss": 0.0,
+      "step": 17770
+    },
+    {
+      "epoch": 75.00982905982906,
+      "grad_norm": 0.0001800288155209273,
+      "learning_rate": 2.6685660018993355e-06,
+      "loss": 0.6123,
+      "step": 17780
+    },
+    {
+      "epoch": 75.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.478565216064453,
+      "eval_runtime": 33.5224,
+      "eval_samples_per_second": 0.746,
+      "eval_steps_per_second": 0.746,
+      "step": 17784
+    },
+    {
+      "epoch": 76.00025641025641,
+      "grad_norm": 0.00023036974016577005,
+      "learning_rate": 2.663817663817664e-06,
+      "loss": 0.0,
+      "step": 17790
+    },
+    {
+      "epoch": 76.00068376068376,
+      "grad_norm": 0.0002549351193010807,
+      "learning_rate": 2.659069325735993e-06,
+      "loss": 0.0,
+      "step": 17800
+    },
+    {
+      "epoch": 76.00111111111111,
+      "grad_norm": 0.00015945105405990034,
+      "learning_rate": 2.6543209876543212e-06,
+      "loss": 0.0768,
+      "step": 17810
+    },
+    {
+      "epoch": 76.00153846153846,
+      "grad_norm": 0.00026457657804712653,
+      "learning_rate": 2.64957264957265e-06,
+      "loss": 0.0007,
+      "step": 17820
+    },
+    {
+      "epoch": 76.00196581196582,
+      "grad_norm": 0.0002729761472437531,
+      "learning_rate": 2.6448243114909783e-06,
+      "loss": 0.0,
+      "step": 17830
+    },
+    {
+      "epoch": 76.00239316239316,
+      "grad_norm": 0.00020036596106365323,
+      "learning_rate": 2.6400759734093066e-06,
+      "loss": 0.0,
+      "step": 17840
+    },
+    {
+      "epoch": 76.0028205128205,
+      "grad_norm": 0.00016762949235271662,
+      "learning_rate": 2.6353276353276357e-06,
+      "loss": 0.0,
+      "step": 17850
+    },
+    {
+      "epoch": 76.00324786324786,
+      "grad_norm": 0.0004586647264659405,
+      "learning_rate": 2.6305792972459644e-06,
+      "loss": 0.0,
+      "step": 17860
+    },
+    {
+      "epoch": 76.00367521367521,
+      "grad_norm": 0.006451745051890612,
+      "learning_rate": 2.6258309591642927e-06,
+      "loss": 0.0,
+      "step": 17870
+    },
+    {
+      "epoch": 76.00410256410257,
+      "grad_norm": 0.0031748872715979815,
+      "learning_rate": 2.6210826210826214e-06,
+      "loss": 0.0,
+      "step": 17880
+    },
+    {
+      "epoch": 76.00452991452991,
+      "grad_norm": 0.0012073371326550841,
+      "learning_rate": 2.6163342830009497e-06,
+      "loss": 0.0,
+      "step": 17890
+    },
+    {
+      "epoch": 76.00495726495727,
+      "grad_norm": 0.0001654605002840981,
+      "learning_rate": 2.611585944919278e-06,
+      "loss": 0.0,
+      "step": 17900
+    },
+    {
+      "epoch": 76.00538461538461,
+      "grad_norm": 0.0012388996547088027,
+      "learning_rate": 2.606837606837607e-06,
+      "loss": 0.0,
+      "step": 17910
+    },
+    {
+      "epoch": 76.00581196581197,
+      "grad_norm": 0.00017125860904343426,
+      "learning_rate": 2.602089268755936e-06,
+      "loss": 0.0,
+      "step": 17920
+    },
+    {
+      "epoch": 76.00623931623932,
+      "grad_norm": 0.004054773598909378,
+      "learning_rate": 2.597340930674264e-06,
+      "loss": 0.0001,
+      "step": 17930
+    },
+    {
+      "epoch": 76.00666666666666,
+      "grad_norm": 0.0019136080518364906,
+      "learning_rate": 2.5925925925925925e-06,
+      "loss": 0.0,
+      "step": 17940
+    },
+    {
+      "epoch": 76.00709401709402,
+      "grad_norm": 0.0007586510619148612,
+      "learning_rate": 2.587844254510921e-06,
+      "loss": 0.0,
+      "step": 17950
+    },
+    {
+      "epoch": 76.00752136752136,
+      "grad_norm": 0.0010209938045591116,
+      "learning_rate": 2.5830959164292504e-06,
+      "loss": 0.0004,
+      "step": 17960
+    },
+    {
+      "epoch": 76.00794871794872,
+      "grad_norm": 0.00018688829732127488,
+      "learning_rate": 2.5783475783475787e-06,
+      "loss": 0.0,
+      "step": 17970
+    },
+    {
+      "epoch": 76.00837606837607,
+      "grad_norm": 0.00018369669851381332,
+      "learning_rate": 2.573599240265907e-06,
+      "loss": 0.0,
+      "step": 17980
+    },
+    {
+      "epoch": 76.00880341880342,
+      "grad_norm": 0.0010952269658446312,
+      "learning_rate": 2.5688509021842357e-06,
+      "loss": 0.0,
+      "step": 17990
+    },
+    {
+      "epoch": 76.00923076923077,
+      "grad_norm": 0.00021538576402235776,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 0.0003,
+      "step": 18000
+    },
+    {
+      "epoch": 76.00965811965811,
+      "grad_norm": 0.0005051797488704324,
+      "learning_rate": 2.559354226020893e-06,
+      "loss": 0.0,
+      "step": 18010
+    },
+    {
+      "epoch": 76.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.552752494812012,
+      "eval_runtime": 33.6617,
+      "eval_samples_per_second": 0.743,
+      "eval_steps_per_second": 0.743,
+      "step": 18018
+    },
+    {
+      "epoch": 77.00008547008547,
+      "grad_norm": 0.000823773501906544,
+      "learning_rate": 2.5546058879392214e-06,
+      "loss": 0.0004,
+      "step": 18020
+    },
+    {
+      "epoch": 77.00051282051282,
+      "grad_norm": 0.001240988029167056,
+      "learning_rate": 2.54985754985755e-06,
+      "loss": 0.0,
+      "step": 18030
+    },
+    {
+      "epoch": 77.00094017094017,
+      "grad_norm": 0.0006408431218005717,
+      "learning_rate": 2.5451092117758784e-06,
+      "loss": 0.0001,
+      "step": 18040
+    },
+    {
+      "epoch": 77.00136752136753,
+      "grad_norm": 0.00021658137848135084,
+      "learning_rate": 2.540360873694207e-06,
+      "loss": 0.0,
+      "step": 18050
+    },
+    {
+      "epoch": 77.00179487179487,
+      "grad_norm": 0.00019429672101978213,
+      "learning_rate": 2.535612535612536e-06,
+      "loss": 0.0,
+      "step": 18060
+    },
+    {
+      "epoch": 77.00222222222222,
+      "grad_norm": 0.00022740900749340653,
+      "learning_rate": 2.5308641975308646e-06,
+      "loss": 0.0,
+      "step": 18070
+    },
+    {
+      "epoch": 77.00264957264957,
+      "grad_norm": 0.00015916908159852028,
+      "learning_rate": 2.526115859449193e-06,
+      "loss": 0.7221,
+      "step": 18080
+    },
+    {
+      "epoch": 77.00307692307692,
+      "grad_norm": 0.00014186595217324793,
+      "learning_rate": 2.5213675213675216e-06,
+      "loss": 0.0,
+      "step": 18090
+    },
+    {
+      "epoch": 77.00350427350428,
+      "grad_norm": 0.00016282236902043223,
+      "learning_rate": 2.51661918328585e-06,
+      "loss": 0.0,
+      "step": 18100
+    },
+    {
+      "epoch": 77.00393162393162,
+      "grad_norm": 0.0003063238982576877,
+      "learning_rate": 2.511870845204179e-06,
+      "loss": 0.0,
+      "step": 18110
+    },
+    {
+      "epoch": 77.00435897435898,
+      "grad_norm": 0.00018556939903646708,
+      "learning_rate": 2.5071225071225073e-06,
+      "loss": 0.0,
+      "step": 18120
+    },
+    {
+      "epoch": 77.00478632478632,
+      "grad_norm": 0.00014702802582178265,
+      "learning_rate": 2.502374169040836e-06,
+      "loss": 0.0,
+      "step": 18130
+    },
+    {
+      "epoch": 77.00521367521368,
+      "grad_norm": 0.004362730775028467,
+      "learning_rate": 2.4976258309591644e-06,
+      "loss": 0.0,
+      "step": 18140
+    },
+    {
+      "epoch": 77.00564102564103,
+      "grad_norm": 0.00015520227316301316,
+      "learning_rate": 2.492877492877493e-06,
+      "loss": 0.0013,
+      "step": 18150
+    },
+    {
+      "epoch": 77.00606837606837,
+      "grad_norm": 528.2244262695312,
+      "learning_rate": 2.488129154795822e-06,
+      "loss": 0.274,
+      "step": 18160
+    },
+    {
+      "epoch": 77.00649572649573,
+      "grad_norm": 0.00016038109606597573,
+      "learning_rate": 2.48338081671415e-06,
+      "loss": 0.0,
+      "step": 18170
+    },
+    {
+      "epoch": 77.00692307692307,
+      "grad_norm": 0.0002260198671137914,
+      "learning_rate": 2.478632478632479e-06,
+      "loss": 0.0022,
+      "step": 18180
+    },
+    {
+      "epoch": 77.00735042735043,
+      "grad_norm": 0.008054505102336407,
+      "learning_rate": 2.4738841405508075e-06,
+      "loss": 0.0718,
+      "step": 18190
+    },
+    {
+      "epoch": 77.00777777777778,
+      "grad_norm": 0.0019935595337301493,
+      "learning_rate": 2.469135802469136e-06,
+      "loss": 0.0,
+      "step": 18200
+    },
+    {
+      "epoch": 77.00820512820513,
+      "grad_norm": 0.0007075904286466539,
+      "learning_rate": 2.4643874643874645e-06,
+      "loss": 0.0,
+      "step": 18210
+    },
+    {
+      "epoch": 77.00863247863248,
+      "grad_norm": 0.0013958788476884365,
+      "learning_rate": 2.4596391263057933e-06,
+      "loss": 0.0,
+      "step": 18220
+    },
+    {
+      "epoch": 77.00905982905982,
+      "grad_norm": 0.0016203945269808173,
+      "learning_rate": 2.454890788224122e-06,
+      "loss": 0.0,
+      "step": 18230
+    },
+    {
+      "epoch": 77.00948717948718,
+      "grad_norm": 0.00014267765800468624,
+      "learning_rate": 2.4501424501424503e-06,
+      "loss": 0.0,
+      "step": 18240
+    },
+    {
+      "epoch": 77.00991452991452,
+      "grad_norm": 0.001626075361855328,
+      "learning_rate": 2.445394112060779e-06,
+      "loss": 0.0,
+      "step": 18250
+    },
+    {
+      "epoch": 77.01,
+      "eval_accuracy": 0.44,
+      "eval_loss": 5.542598724365234,
+      "eval_runtime": 33.0633,
+      "eval_samples_per_second": 0.756,
+      "eval_steps_per_second": 0.756,
+      "step": 18252
+    },
+    {
+      "epoch": 78.00034188034188,
+      "grad_norm": 0.0003114393330179155,
+      "learning_rate": 2.4406457739791077e-06,
+      "loss": 0.0,
+      "step": 18260
+    },
+    {
+      "epoch": 78.00076923076924,
+      "grad_norm": 0.00026152783539146185,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 0.0,
+      "step": 18270
+    },
+    {
+      "epoch": 78.00119658119658,
+      "grad_norm": 0.0003940732858609408,
+      "learning_rate": 2.4311490978157647e-06,
+      "loss": 0.0,
+      "step": 18280
+    },
+    {
+      "epoch": 78.00162393162393,
+      "grad_norm": 0.000839279149658978,
+      "learning_rate": 2.4264007597340935e-06,
+      "loss": 0.0059,
+      "step": 18290
+    },
+    {
+      "epoch": 78.00205128205128,
+      "grad_norm": 0.00048554647946730256,
+      "learning_rate": 2.4216524216524218e-06,
+      "loss": 0.0,
+      "step": 18300
+    },
+    {
+      "epoch": 78.00247863247863,
+      "grad_norm": 0.00023615985992364585,
+      "learning_rate": 2.4169040835707505e-06,
+      "loss": 0.0,
+      "step": 18310
+    },
+    {
+      "epoch": 78.00290598290599,
+      "grad_norm": 0.0001385568466503173,
+      "learning_rate": 2.412155745489079e-06,
+      "loss": 0.0,
+      "step": 18320
+    },
+    {
+      "epoch": 78.00333333333333,
+      "grad_norm": 0.0001412118726875633,
+      "learning_rate": 2.4074074074074075e-06,
+      "loss": 0.0,
+      "step": 18330
+    },
+    {
+      "epoch": 78.00376068376069,
+      "grad_norm": 0.0002054434735327959,
+      "learning_rate": 2.4026590693257362e-06,
+      "loss": 0.0,
+      "step": 18340
+    },
+    {
+      "epoch": 78.00418803418803,
+      "grad_norm": 0.0019602985121309757,
+      "learning_rate": 2.3979107312440645e-06,
+      "loss": 0.0,
+      "step": 18350
+    },
+    {
+      "epoch": 78.00461538461539,
+      "grad_norm": 0.00021056836703792214,
+      "learning_rate": 2.3931623931623937e-06,
+      "loss": 0.0,
+      "step": 18360
+    },
+    {
+      "epoch": 78.00504273504274,
+      "grad_norm": 40.98805618286133,
+      "learning_rate": 2.388414055080722e-06,
+      "loss": 0.0028,
+      "step": 18370
+    },
+    {
+      "epoch": 78.00547008547008,
+      "grad_norm": 0.00020283031335566193,
+      "learning_rate": 2.3836657169990502e-06,
+      "loss": 0.0,
+      "step": 18380
+    },
+    {
+      "epoch": 78.00589743589744,
+      "grad_norm": 0.00025577127235010266,
+      "learning_rate": 2.378917378917379e-06,
+      "loss": 0.0001,
+      "step": 18390
+    },
+    {
+      "epoch": 78.00632478632478,
+      "grad_norm": 0.00014881067909300327,
+      "learning_rate": 2.3741690408357077e-06,
+      "loss": 0.0002,
+      "step": 18400
+    },
+    {
+      "epoch": 78.00675213675214,
+      "grad_norm": 0.00021390440815594047,
+      "learning_rate": 2.3694207027540364e-06,
+      "loss": 0.0,
+      "step": 18410
+    },
+    {
+      "epoch": 78.00717948717949,
+      "grad_norm": 0.00022567427367903292,
+      "learning_rate": 2.3646723646723647e-06,
+      "loss": 0.0,
+      "step": 18420
+    },
+    {
+      "epoch": 78.00760683760684,
+      "grad_norm": 0.00015315061318688095,
+      "learning_rate": 2.3599240265906934e-06,
+      "loss": 0.0,
+      "step": 18430
+    },
+    {
+      "epoch": 78.00803418803419,
+      "grad_norm": 0.00031218226649798453,
+      "learning_rate": 2.355175688509022e-06,
+      "loss": 0.0009,
+      "step": 18440
+    },
+    {
+      "epoch": 78.00846153846153,
+      "grad_norm": 0.6312761902809143,
+      "learning_rate": 2.3504273504273504e-06,
+      "loss": 0.0002,
+      "step": 18450
+    },
+    {
+      "epoch": 78.00888888888889,
+      "grad_norm": 0.13864165544509888,
+      "learning_rate": 2.345679012345679e-06,
+      "loss": 0.0003,
+      "step": 18460
+    },
+    {
+      "epoch": 78.00931623931623,
+      "grad_norm": 0.0004563250986393541,
+      "learning_rate": 2.340930674264008e-06,
+      "loss": 0.0006,
+      "step": 18470
+    },
+    {
+      "epoch": 78.0097435897436,
+      "grad_norm": 0.0005144558963365853,
+      "learning_rate": 2.336182336182336e-06,
+      "loss": 0.0,
+      "step": 18480
+    },
+    {
+      "epoch": 78.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.4275970458984375,
+      "eval_runtime": 33.5576,
+      "eval_samples_per_second": 0.745,
+      "eval_steps_per_second": 0.745,
+      "step": 18486
+    },
+    {
+      "epoch": 79.00017094017095,
+      "grad_norm": 0.00022311207430902869,
+      "learning_rate": 2.331433998100665e-06,
+      "loss": 0.0038,
+      "step": 18490
+    },
+    {
+      "epoch": 79.00059829059829,
+      "grad_norm": 0.0002333286392968148,
+      "learning_rate": 2.3266856600189936e-06,
+      "loss": 0.0,
+      "step": 18500
+    },
+    {
+      "epoch": 79.00102564102563,
+      "grad_norm": 0.00024575545103289187,
+      "learning_rate": 2.321937321937322e-06,
+      "loss": 0.0001,
+      "step": 18510
+    },
+    {
+      "epoch": 79.001452991453,
+      "grad_norm": 0.0004118859360460192,
+      "learning_rate": 2.3171889838556506e-06,
+      "loss": 0.0,
+      "step": 18520
+    },
+    {
+      "epoch": 79.00188034188034,
+      "grad_norm": 0.0018699835054576397,
+      "learning_rate": 2.3124406457739794e-06,
+      "loss": 0.0,
+      "step": 18530
+    },
+    {
+      "epoch": 79.0023076923077,
+      "grad_norm": 0.0005015349015593529,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 0.0,
+      "step": 18540
+    },
+    {
+      "epoch": 79.00273504273504,
+      "grad_norm": 0.00027656828751787543,
+      "learning_rate": 2.3029439696106364e-06,
+      "loss": 0.0,
+      "step": 18550
+    },
+    {
+      "epoch": 79.0031623931624,
+      "grad_norm": 0.0010210457257926464,
+      "learning_rate": 2.298195631528965e-06,
+      "loss": 0.0,
+      "step": 18560
+    },
+    {
+      "epoch": 79.00358974358974,
+      "grad_norm": 0.0015956490533426404,
+      "learning_rate": 2.293447293447294e-06,
+      "loss": 0.0,
+      "step": 18570
+    },
+    {
+      "epoch": 79.0040170940171,
+      "grad_norm": 0.001694871811196208,
+      "learning_rate": 2.288698955365622e-06,
+      "loss": 0.0,
+      "step": 18580
+    },
+    {
+      "epoch": 79.00444444444445,
+      "grad_norm": 0.0004087260167580098,
+      "learning_rate": 2.283950617283951e-06,
+      "loss": 0.0,
+      "step": 18590
+    },
+    {
+      "epoch": 79.00487179487179,
+      "grad_norm": 0.00020924824639223516,
+      "learning_rate": 2.2792022792022796e-06,
+      "loss": 0.0,
+      "step": 18600
+    },
+    {
+      "epoch": 79.00529914529915,
+      "grad_norm": 0.0006689398433081806,
+      "learning_rate": 2.274453941120608e-06,
+      "loss": 0.0,
+      "step": 18610
+    },
+    {
+      "epoch": 79.00572649572649,
+      "grad_norm": 0.0007431868580169976,
+      "learning_rate": 2.2697056030389366e-06,
+      "loss": 0.5593,
+      "step": 18620
+    },
+    {
+      "epoch": 79.00615384615385,
+      "grad_norm": 0.00014568243932444602,
+      "learning_rate": 2.2649572649572653e-06,
+      "loss": 0.0001,
+      "step": 18630
+    },
+    {
+      "epoch": 79.0065811965812,
+      "grad_norm": 0.0003919856681022793,
+      "learning_rate": 2.2602089268755936e-06,
+      "loss": 0.0,
+      "step": 18640
+    },
+    {
+      "epoch": 79.00700854700855,
+      "grad_norm": 0.00014353814185597003,
+      "learning_rate": 2.2554605887939223e-06,
+      "loss": 0.0,
+      "step": 18650
+    },
+    {
+      "epoch": 79.0074358974359,
+      "grad_norm": 0.00022350714425556362,
+      "learning_rate": 2.250712250712251e-06,
+      "loss": 0.0,
+      "step": 18660
+    },
+    {
+      "epoch": 79.00786324786324,
+      "grad_norm": 0.0012620558263733983,
+      "learning_rate": 2.2459639126305797e-06,
+      "loss": 0.0012,
+      "step": 18670
+    },
+    {
+      "epoch": 79.0082905982906,
+      "grad_norm": 0.0014929536264389753,
+      "learning_rate": 2.241215574548908e-06,
+      "loss": 0.0,
+      "step": 18680
+    },
+    {
+      "epoch": 79.00871794871794,
+      "grad_norm": 0.0001310633379034698,
+      "learning_rate": 2.2364672364672368e-06,
+      "loss": 0.0,
+      "step": 18690
+    },
+    {
+      "epoch": 79.0091452991453,
+      "grad_norm": 2.5448873043060303,
+      "learning_rate": 2.2317188983855655e-06,
+      "loss": 0.0006,
+      "step": 18700
+    },
+    {
+      "epoch": 79.00957264957265,
+      "grad_norm": 0.0006994788418523967,
+      "learning_rate": 2.2269705603038938e-06,
+      "loss": 0.0,
+      "step": 18710
+    },
+    {
+      "epoch": 79.01,
+      "grad_norm": 0.00040815817192196846,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 0.0,
+      "step": 18720
+    },
+    {
+      "epoch": 79.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.867609024047852,
+      "eval_runtime": 33.9385,
+      "eval_samples_per_second": 0.737,
+      "eval_steps_per_second": 0.737,
+      "step": 18720
+    },
+    {
+      "epoch": 80.00042735042734,
+      "grad_norm": 0.0001611685729585588,
+      "learning_rate": 2.2174738841405512e-06,
+      "loss": 0.0,
+      "step": 18730
+    },
+    {
+      "epoch": 80.0008547008547,
+      "grad_norm": 0.00016659338143654168,
+      "learning_rate": 2.2127255460588795e-06,
+      "loss": 0.0,
+      "step": 18740
+    },
+    {
+      "epoch": 80.00128205128205,
+      "grad_norm": 0.0005343634402379394,
+      "learning_rate": 2.2079772079772082e-06,
+      "loss": 0.0,
+      "step": 18750
+    },
+    {
+      "epoch": 80.0017094017094,
+      "grad_norm": 0.00014189437206368893,
+      "learning_rate": 2.203228869895537e-06,
+      "loss": 0.0,
+      "step": 18760
+    },
+    {
+      "epoch": 80.00213675213675,
+      "grad_norm": 0.00016400543972849846,
+      "learning_rate": 2.1984805318138653e-06,
+      "loss": 0.0,
+      "step": 18770
+    },
+    {
+      "epoch": 80.00256410256411,
+      "grad_norm": 0.00013671614578925073,
+      "learning_rate": 2.193732193732194e-06,
+      "loss": 0.0,
+      "step": 18780
+    },
+    {
+      "epoch": 80.00299145299145,
+      "grad_norm": 0.0015447050100192428,
+      "learning_rate": 2.1889838556505223e-06,
+      "loss": 0.0,
+      "step": 18790
+    },
+    {
+      "epoch": 80.00341880341881,
+      "grad_norm": 0.00014961874694563448,
+      "learning_rate": 2.1842355175688514e-06,
+      "loss": 0.0,
+      "step": 18800
+    },
+    {
+      "epoch": 80.00384615384615,
+      "grad_norm": 0.000166281548445113,
+      "learning_rate": 2.1794871794871797e-06,
+      "loss": 0.2076,
+      "step": 18810
+    },
+    {
+      "epoch": 80.0042735042735,
+      "grad_norm": 0.0008461292018182576,
+      "learning_rate": 2.174738841405508e-06,
+      "loss": 0.0,
+      "step": 18820
+    },
+    {
+      "epoch": 80.00470085470086,
+      "grad_norm": 0.00018119644664693624,
+      "learning_rate": 2.1699905033238367e-06,
+      "loss": 1.0097,
+      "step": 18830
+    },
+    {
+      "epoch": 80.0051282051282,
+      "grad_norm": 0.0001796925498638302,
+      "learning_rate": 2.1652421652421654e-06,
+      "loss": 0.0,
+      "step": 18840
+    },
+    {
+      "epoch": 80.00555555555556,
+      "grad_norm": 0.0001590051979292184,
+      "learning_rate": 2.1604938271604937e-06,
+      "loss": 0.0917,
+      "step": 18850
+    },
+    {
+      "epoch": 80.0059829059829,
+      "grad_norm": 0.00016695998783688992,
+      "learning_rate": 2.1557454890788225e-06,
+      "loss": 0.0,
+      "step": 18860
+    },
+    {
+      "epoch": 80.00641025641026,
+      "grad_norm": 0.0002668647503014654,
+      "learning_rate": 2.150997150997151e-06,
+      "loss": 0.0,
+      "step": 18870
+    },
+    {
+      "epoch": 80.0068376068376,
+      "grad_norm": 0.0016600067028775811,
+      "learning_rate": 2.14624881291548e-06,
+      "loss": 0.0,
+      "step": 18880
+    },
+    {
+      "epoch": 80.00726495726495,
+      "grad_norm": 0.0004276032268535346,
+      "learning_rate": 2.141500474833808e-06,
+      "loss": 0.0,
+      "step": 18890
+    },
+    {
+      "epoch": 80.00769230769231,
+      "grad_norm": 0.5101644992828369,
+      "learning_rate": 2.136752136752137e-06,
+      "loss": 0.0001,
+      "step": 18900
+    },
+    {
+      "epoch": 80.00811965811965,
+      "grad_norm": 0.00014222673780750483,
+      "learning_rate": 2.1320037986704656e-06,
+      "loss": 0.0,
+      "step": 18910
+    },
+    {
+      "epoch": 80.00854700854701,
+      "grad_norm": 0.0010868561221286654,
+      "learning_rate": 2.127255460588794e-06,
+      "loss": 0.0,
+      "step": 18920
+    },
+    {
+      "epoch": 80.00897435897436,
+      "grad_norm": 0.0014827632112428546,
+      "learning_rate": 2.1225071225071227e-06,
+      "loss": 0.0,
+      "step": 18930
+    },
+    {
+      "epoch": 80.00940170940171,
+      "grad_norm": 0.00015382345009129494,
+      "learning_rate": 2.1177587844254514e-06,
+      "loss": 0.0,
+      "step": 18940
+    },
+    {
+      "epoch": 80.00982905982906,
+      "grad_norm": 0.00015482779417652637,
+      "learning_rate": 2.1130104463437797e-06,
+      "loss": 0.0,
+      "step": 18950
+    },
+    {
+      "epoch": 80.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.669287204742432,
+      "eval_runtime": 33.799,
+      "eval_samples_per_second": 0.74,
+      "eval_steps_per_second": 0.74,
+      "step": 18954
+    },
+    {
+      "epoch": 81.00025641025641,
+      "grad_norm": 0.00021314578771125525,
+      "learning_rate": 2.1082621082621084e-06,
+      "loss": 0.0002,
+      "step": 18960
+    },
+    {
+      "epoch": 81.00068376068376,
+      "grad_norm": 0.00015724689001217484,
+      "learning_rate": 2.103513770180437e-06,
+      "loss": 0.0,
+      "step": 18970
+    },
+    {
+      "epoch": 81.00111111111111,
+      "grad_norm": 0.00015798299864400178,
+      "learning_rate": 2.0987654320987654e-06,
+      "loss": 0.6005,
+      "step": 18980
+    },
+    {
+      "epoch": 81.00153846153846,
+      "grad_norm": 0.00015623167564626783,
+      "learning_rate": 2.094017094017094e-06,
+      "loss": 0.0,
+      "step": 18990
+    },
+    {
+      "epoch": 81.00196581196582,
+      "grad_norm": 0.001082591712474823,
+      "learning_rate": 2.089268755935423e-06,
+      "loss": 0.0,
+      "step": 19000
+    },
+    {
+      "epoch": 81.00239316239316,
+      "grad_norm": 0.00016879322356544435,
+      "learning_rate": 2.0845204178537516e-06,
+      "loss": 0.0001,
+      "step": 19010
+    },
+    {
+      "epoch": 81.0028205128205,
+      "grad_norm": 0.0006452444358728826,
+      "learning_rate": 2.07977207977208e-06,
+      "loss": 0.0,
+      "step": 19020
+    },
+    {
+      "epoch": 81.00324786324786,
+      "grad_norm": 0.00027353313635103405,
+      "learning_rate": 2.0750237416904086e-06,
+      "loss": 0.0,
+      "step": 19030
+    },
+    {
+      "epoch": 81.00367521367521,
+      "grad_norm": 0.0008052527555264533,
+      "learning_rate": 2.0702754036087373e-06,
+      "loss": 0.0,
+      "step": 19040
+    },
+    {
+      "epoch": 81.00410256410257,
+      "grad_norm": 0.00014285088400356472,
+      "learning_rate": 2.0655270655270656e-06,
+      "loss": 0.0,
+      "step": 19050
+    },
+    {
+      "epoch": 81.00452991452991,
+      "grad_norm": 0.00035150200710631907,
+      "learning_rate": 2.0607787274453943e-06,
+      "loss": 0.0013,
+      "step": 19060
+    },
+    {
+      "epoch": 81.00495726495727,
+      "grad_norm": 0.0010856334120035172,
+      "learning_rate": 2.056030389363723e-06,
+      "loss": 0.0,
+      "step": 19070
+    },
+    {
+      "epoch": 81.00538461538461,
+      "grad_norm": 0.0009435404208488762,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 0.0,
+      "step": 19080
+    },
+    {
+      "epoch": 81.00581196581197,
+      "grad_norm": 0.001047900877892971,
+      "learning_rate": 2.04653371320038e-06,
+      "loss": 0.0,
+      "step": 19090
+    },
+    {
+      "epoch": 81.00623931623932,
+      "grad_norm": 0.00044585313298739493,
+      "learning_rate": 2.0417853751187088e-06,
+      "loss": 0.0,
+      "step": 19100
+    },
+    {
+      "epoch": 81.00666666666666,
+      "grad_norm": 0.00014062630361877382,
+      "learning_rate": 2.037037037037037e-06,
+      "loss": 0.0,
+      "step": 19110
+    },
+    {
+      "epoch": 81.00709401709402,
+      "grad_norm": 0.00024204060900956392,
+      "learning_rate": 2.032288698955366e-06,
+      "loss": 0.0,
+      "step": 19120
+    },
+    {
+      "epoch": 81.00752136752136,
+      "grad_norm": 0.000514078012201935,
+      "learning_rate": 2.0275403608736945e-06,
+      "loss": 0.0128,
+      "step": 19130
+    },
+    {
+      "epoch": 81.00794871794872,
+      "grad_norm": 0.0003828479675576091,
+      "learning_rate": 2.0227920227920232e-06,
+      "loss": 0.0,
+      "step": 19140
+    },
+    {
+      "epoch": 81.00837606837607,
+      "grad_norm": 0.000210257523576729,
+      "learning_rate": 2.0180436847103515e-06,
+      "loss": 0.0,
+      "step": 19150
+    },
+    {
+      "epoch": 81.00880341880342,
+      "grad_norm": 0.00039357185596600175,
+      "learning_rate": 2.01329534662868e-06,
+      "loss": 0.0,
+      "step": 19160
+    },
+    {
+      "epoch": 81.00923076923077,
+      "grad_norm": 0.00014624725736211985,
+      "learning_rate": 2.008547008547009e-06,
+      "loss": 0.0,
+      "step": 19170
+    },
+    {
+      "epoch": 81.00965811965811,
+      "grad_norm": 0.0019884561188519,
+      "learning_rate": 2.0037986704653373e-06,
+      "loss": 0.0,
+      "step": 19180
+    },
+    {
+      "epoch": 81.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.791877269744873,
+      "eval_runtime": 33.6676,
+      "eval_samples_per_second": 0.743,
+      "eval_steps_per_second": 0.743,
+      "step": 19188
+    },
+    {
+      "epoch": 82.00008547008547,
+      "grad_norm": 0.00014335830928757787,
+      "learning_rate": 1.9990503323836656e-06,
+      "loss": 0.0,
+      "step": 19190
+    },
+    {
+      "epoch": 82.00051282051282,
+      "grad_norm": 0.00016715576930437237,
+      "learning_rate": 1.9943019943019947e-06,
+      "loss": 0.0,
+      "step": 19200
+    },
+    {
+      "epoch": 82.00094017094017,
+      "grad_norm": 0.00015401339624077082,
+      "learning_rate": 1.989553656220323e-06,
+      "loss": 0.0,
+      "step": 19210
+    },
+    {
+      "epoch": 82.00136752136753,
+      "grad_norm": 0.0009122826741077006,
+      "learning_rate": 1.9848053181386517e-06,
+      "loss": 0.0,
+      "step": 19220
+    },
+    {
+      "epoch": 82.00179487179487,
+      "grad_norm": 0.0039600650779902935,
+      "learning_rate": 1.98005698005698e-06,
+      "loss": 0.0,
+      "step": 19230
+    },
+    {
+      "epoch": 82.00222222222222,
+      "grad_norm": 0.00023766346566844732,
+      "learning_rate": 1.9753086419753087e-06,
+      "loss": 0.0,
+      "step": 19240
+    },
+    {
+      "epoch": 82.00264957264957,
+      "grad_norm": 0.0002676662988960743,
+      "learning_rate": 1.9705603038936375e-06,
+      "loss": 0.0,
+      "step": 19250
+    },
+    {
+      "epoch": 82.00307692307692,
+      "grad_norm": 0.00012970353418495506,
+      "learning_rate": 1.9658119658119658e-06,
+      "loss": 0.9915,
+      "step": 19260
+    },
+    {
+      "epoch": 82.00350427350428,
+      "grad_norm": 0.001931848586536944,
+      "learning_rate": 1.9610636277302945e-06,
+      "loss": 0.0,
+      "step": 19270
+    },
+    {
+      "epoch": 82.00393162393162,
+      "grad_norm": 0.00016302384028676897,
+      "learning_rate": 1.956315289648623e-06,
+      "loss": 0.0,
+      "step": 19280
+    },
+    {
+      "epoch": 82.00435897435898,
+      "grad_norm": 0.0006244336836971343,
+      "learning_rate": 1.9515669515669515e-06,
+      "loss": 0.0,
+      "step": 19290
+    },
+    {
+      "epoch": 82.00478632478632,
+      "grad_norm": 0.0013233753852546215,
+      "learning_rate": 1.9468186134852802e-06,
+      "loss": 0.0,
+      "step": 19300
+    },
+    {
+      "epoch": 82.00521367521368,
+      "grad_norm": 0.00016536538896616548,
+      "learning_rate": 1.942070275403609e-06,
+      "loss": 0.0,
+      "step": 19310
+    },
+    {
+      "epoch": 82.00564102564103,
+      "grad_norm": 0.002657419303432107,
+      "learning_rate": 1.9373219373219372e-06,
+      "loss": 0.0276,
+      "step": 19320
+    },
+    {
+      "epoch": 82.00606837606837,
+      "grad_norm": 0.000514014158397913,
+      "learning_rate": 1.932573599240266e-06,
+      "loss": 0.0,
+      "step": 19330
+    },
+    {
+      "epoch": 82.00649572649573,
+      "grad_norm": 0.000511766062118113,
+      "learning_rate": 1.9278252611585947e-06,
+      "loss": 0.0,
+      "step": 19340
+    },
+    {
+      "epoch": 82.00692307692307,
+      "grad_norm": 0.0024127494543790817,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 0.0,
+      "step": 19350
+    },
+    {
+      "epoch": 82.00735042735043,
+      "grad_norm": 0.0032117695081979036,
+      "learning_rate": 1.9183285849952517e-06,
+      "loss": 0.0,
+      "step": 19360
+    },
+    {
+      "epoch": 82.00777777777778,
+      "grad_norm": 0.0002099874400300905,
+      "learning_rate": 1.9135802469135804e-06,
+      "loss": 0.0,
+      "step": 19370
+    },
+    {
+      "epoch": 82.00820512820513,
+      "grad_norm": 0.0007865933002904058,
+      "learning_rate": 1.908831908831909e-06,
+      "loss": 0.0,
+      "step": 19380
+    },
+    {
+      "epoch": 82.00863247863248,
+      "grad_norm": 0.000656930438708514,
+      "learning_rate": 1.9040835707502374e-06,
+      "loss": 0.0,
+      "step": 19390
+    },
+    {
+      "epoch": 82.00905982905982,
+      "grad_norm": 0.16030727326869965,
+      "learning_rate": 1.8993352326685664e-06,
+      "loss": 0.0001,
+      "step": 19400
+    },
+    {
+      "epoch": 82.00948717948718,
+      "grad_norm": 0.0001436812017345801,
+      "learning_rate": 1.8945868945868947e-06,
+      "loss": 0.0,
+      "step": 19410
+    },
+    {
+      "epoch": 82.00991452991452,
+      "grad_norm": 0.000723684614058584,
+      "learning_rate": 1.8898385565052232e-06,
+      "loss": 0.0,
+      "step": 19420
+    },
+    {
+      "epoch": 82.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.751951694488525,
+      "eval_runtime": 36.0289,
+      "eval_samples_per_second": 0.694,
+      "eval_steps_per_second": 0.694,
+      "step": 19422
+    },
+    {
+      "epoch": 83.00034188034188,
+      "grad_norm": 0.000560764514375478,
+      "learning_rate": 1.885090218423552e-06,
+      "loss": 0.0,
+      "step": 19430
+    },
+    {
+      "epoch": 83.00076923076924,
+      "grad_norm": 0.00022796269331593066,
+      "learning_rate": 1.8803418803418804e-06,
+      "loss": 0.0,
+      "step": 19440
+    },
+    {
+      "epoch": 83.00119658119658,
+      "grad_norm": 0.00013079910422675312,
+      "learning_rate": 1.875593542260209e-06,
+      "loss": 0.0,
+      "step": 19450
+    },
+    {
+      "epoch": 83.00162393162393,
+      "grad_norm": 0.0005907687591388822,
+      "learning_rate": 1.8708452041785376e-06,
+      "loss": 0.3814,
+      "step": 19460
+    },
+    {
+      "epoch": 83.00205128205128,
+      "grad_norm": 0.00020753988064825535,
+      "learning_rate": 1.8660968660968661e-06,
+      "loss": 0.0,
+      "step": 19470
+    },
+    {
+      "epoch": 83.00247863247863,
+      "grad_norm": 0.00022883056954015046,
+      "learning_rate": 1.8613485280151949e-06,
+      "loss": 0.0001,
+      "step": 19480
+    },
+    {
+      "epoch": 83.00290598290599,
+      "grad_norm": 0.0005515534430742264,
+      "learning_rate": 1.8566001899335234e-06,
+      "loss": 0.0003,
+      "step": 19490
+    },
+    {
+      "epoch": 83.00333333333333,
+      "grad_norm": 0.0006961479666642845,
+      "learning_rate": 1.8518518518518519e-06,
+      "loss": 0.0,
+      "step": 19500
+    },
+    {
+      "epoch": 83.00376068376069,
+      "grad_norm": 0.0002969226916320622,
+      "learning_rate": 1.8471035137701806e-06,
+      "loss": 0.0,
+      "step": 19510
+    },
+    {
+      "epoch": 83.00418803418803,
+      "grad_norm": 0.0003522407787386328,
+      "learning_rate": 1.842355175688509e-06,
+      "loss": 0.0,
+      "step": 19520
+    },
+    {
+      "epoch": 83.00461538461539,
+      "grad_norm": 0.0003604775993153453,
+      "learning_rate": 1.8376068376068378e-06,
+      "loss": 0.0,
+      "step": 19530
+    },
+    {
+      "epoch": 83.00504273504274,
+      "grad_norm": 0.00028395134722813964,
+      "learning_rate": 1.8328584995251663e-06,
+      "loss": 0.0022,
+      "step": 19540
+    },
+    {
+      "epoch": 83.00547008547008,
+      "grad_norm": 0.0031416804995387793,
+      "learning_rate": 1.8281101614434948e-06,
+      "loss": 0.0,
+      "step": 19550
+    },
+    {
+      "epoch": 83.00589743589744,
+      "grad_norm": 0.0014400221407413483,
+      "learning_rate": 1.8233618233618236e-06,
+      "loss": 0.0,
+      "step": 19560
+    },
+    {
+      "epoch": 83.00632478632478,
+      "grad_norm": 0.0030254279263317585,
+      "learning_rate": 1.818613485280152e-06,
+      "loss": 0.0,
+      "step": 19570
+    },
+    {
+      "epoch": 83.00675213675214,
+      "grad_norm": 0.00016171699098777026,
+      "learning_rate": 1.8138651471984806e-06,
+      "loss": 0.0,
+      "step": 19580
+    },
+    {
+      "epoch": 83.00717948717949,
+      "grad_norm": 0.3200160264968872,
+      "learning_rate": 1.8091168091168093e-06,
+      "loss": 0.0001,
+      "step": 19590
+    },
+    {
+      "epoch": 83.00760683760684,
+      "grad_norm": 0.00022861655452288687,
+      "learning_rate": 1.8043684710351378e-06,
+      "loss": 0.0006,
+      "step": 19600
+    },
+    {
+      "epoch": 83.00803418803419,
+      "grad_norm": 0.00018940556037705392,
+      "learning_rate": 1.7996201329534665e-06,
+      "loss": 0.0,
+      "step": 19610
+    },
+    {
+      "epoch": 83.00846153846153,
+      "grad_norm": 0.0001795957941794768,
+      "learning_rate": 1.794871794871795e-06,
+      "loss": 0.0,
+      "step": 19620
+    },
+    {
+      "epoch": 83.00888888888889,
+      "grad_norm": 0.00021350740280468017,
+      "learning_rate": 1.7901234567901235e-06,
+      "loss": 0.0,
+      "step": 19630
+    },
+    {
+      "epoch": 83.00931623931623,
+      "grad_norm": 0.0026247838977724314,
+      "learning_rate": 1.7853751187084523e-06,
+      "loss": 0.0,
+      "step": 19640
+    },
+    {
+      "epoch": 83.0097435897436,
+      "grad_norm": 0.0006859219283796847,
+      "learning_rate": 1.7806267806267808e-06,
+      "loss": 0.0,
+      "step": 19650
+    },
+    {
+      "epoch": 83.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.7565107345581055,
+      "eval_runtime": 33.4174,
+      "eval_samples_per_second": 0.748,
+      "eval_steps_per_second": 0.748,
+      "step": 19656
+    },
+    {
+      "epoch": 84.00017094017095,
+      "grad_norm": 0.00018314612680114806,
+      "learning_rate": 1.7758784425451095e-06,
+      "loss": 0.0,
+      "step": 19660
+    },
+    {
+      "epoch": 84.00059829059829,
+      "grad_norm": 0.0016701172571629286,
+      "learning_rate": 1.771130104463438e-06,
+      "loss": 0.0001,
+      "step": 19670
+    },
+    {
+      "epoch": 84.00102564102563,
+      "grad_norm": 0.1983662247657776,
+      "learning_rate": 1.7663817663817665e-06,
+      "loss": 0.0,
+      "step": 19680
+    },
+    {
+      "epoch": 84.001452991453,
+      "grad_norm": 0.00020793025032617152,
+      "learning_rate": 1.7616334283000952e-06,
+      "loss": 0.0,
+      "step": 19690
+    },
+    {
+      "epoch": 84.00188034188034,
+      "grad_norm": 0.00018910918151959777,
+      "learning_rate": 1.7568850902184237e-06,
+      "loss": 0.0,
+      "step": 19700
+    },
+    {
+      "epoch": 84.0023076923077,
+      "grad_norm": 0.0018449919298291206,
+      "learning_rate": 1.7521367521367522e-06,
+      "loss": 0.0,
+      "step": 19710
+    },
+    {
+      "epoch": 84.00273504273504,
+      "grad_norm": 0.0001609017635928467,
+      "learning_rate": 1.747388414055081e-06,
+      "loss": 0.0,
+      "step": 19720
+    },
+    {
+      "epoch": 84.0031623931624,
+      "grad_norm": 0.00035862025106325746,
+      "learning_rate": 1.7426400759734095e-06,
+      "loss": 0.0,
+      "step": 19730
+    },
+    {
+      "epoch": 84.00358974358974,
+      "grad_norm": 0.0008477799710817635,
+      "learning_rate": 1.7378917378917382e-06,
+      "loss": 0.0,
+      "step": 19740
+    },
+    {
+      "epoch": 84.0040170940171,
+      "grad_norm": 0.0038533161859959364,
+      "learning_rate": 1.7331433998100667e-06,
+      "loss": 0.0,
+      "step": 19750
+    },
+    {
+      "epoch": 84.00444444444445,
+      "grad_norm": 0.0002816705673467368,
+      "learning_rate": 1.7283950617283952e-06,
+      "loss": 0.0,
+      "step": 19760
+    },
+    {
+      "epoch": 84.00487179487179,
+      "grad_norm": 0.000693616340868175,
+      "learning_rate": 1.723646723646724e-06,
+      "loss": 0.0,
+      "step": 19770
+    },
+    {
+      "epoch": 84.00529914529915,
+      "grad_norm": 0.00019760453142225742,
+      "learning_rate": 1.7188983855650524e-06,
+      "loss": 0.0001,
+      "step": 19780
+    },
+    {
+      "epoch": 84.00572649572649,
+      "grad_norm": 0.0001434326113667339,
+      "learning_rate": 1.7141500474833807e-06,
+      "loss": 0.0,
+      "step": 19790
+    },
+    {
+      "epoch": 84.00615384615385,
+      "grad_norm": 0.000697402167133987,
+      "learning_rate": 1.7094017094017097e-06,
+      "loss": 0.0,
+      "step": 19800
+    },
+    {
+      "epoch": 84.0065811965812,
+      "grad_norm": 0.0014031692408025265,
+      "learning_rate": 1.704653371320038e-06,
+      "loss": 0.0,
+      "step": 19810
+    },
+    {
+      "epoch": 84.00700854700855,
+      "grad_norm": 0.00014115864178165793,
+      "learning_rate": 1.6999050332383669e-06,
+      "loss": 0.0,
+      "step": 19820
+    },
+    {
+      "epoch": 84.0074358974359,
+      "grad_norm": 0.0013099861098453403,
+      "learning_rate": 1.6951566951566952e-06,
+      "loss": 0.0,
+      "step": 19830
+    },
+    {
+      "epoch": 84.00786324786324,
+      "grad_norm": 0.00024018825206439942,
+      "learning_rate": 1.6904083570750237e-06,
+      "loss": 0.0,
+      "step": 19840
+    },
+    {
+      "epoch": 84.0082905982906,
+      "grad_norm": 0.0007821761537343264,
+      "learning_rate": 1.6856600189933524e-06,
+      "loss": 0.0,
+      "step": 19850
+    },
+    {
+      "epoch": 84.00871794871794,
+      "grad_norm": 0.00013779080472886562,
+      "learning_rate": 1.680911680911681e-06,
+      "loss": 0.0,
+      "step": 19860
+    },
+    {
+      "epoch": 84.0091452991453,
+      "grad_norm": 0.0010995293268933892,
+      "learning_rate": 1.6761633428300099e-06,
+      "loss": 0.0,
+      "step": 19870
+    },
+    {
+      "epoch": 84.00957264957265,
+      "grad_norm": 0.0002249151439173147,
+      "learning_rate": 1.6714150047483382e-06,
+      "loss": 0.0,
+      "step": 19880
+    },
+    {
+      "epoch": 84.01,
+      "grad_norm": 0.00015639988123439252,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.0,
+      "step": 19890
+    },
+    {
+      "epoch": 84.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.818636417388916,
+      "eval_runtime": 33.5438,
+      "eval_samples_per_second": 0.745,
+      "eval_steps_per_second": 0.745,
+      "step": 19890
+    },
+    {
+      "epoch": 85.00042735042734,
+      "grad_norm": 0.0001349115773336962,
+      "learning_rate": 1.6619183285849954e-06,
+      "loss": 0.0,
+      "step": 19900
+    },
+    {
+      "epoch": 85.0008547008547,
+      "grad_norm": 0.00012621526548173279,
+      "learning_rate": 1.6571699905033239e-06,
+      "loss": 0.0,
+      "step": 19910
+    },
+    {
+      "epoch": 85.00128205128205,
+      "grad_norm": 0.00013298312842380255,
+      "learning_rate": 1.6524216524216524e-06,
+      "loss": 0.0,
+      "step": 19920
+    },
+    {
+      "epoch": 85.0017094017094,
+      "grad_norm": 0.005015491507947445,
+      "learning_rate": 1.6476733143399811e-06,
+      "loss": 0.0,
+      "step": 19930
+    },
+    {
+      "epoch": 85.00213675213675,
+      "grad_norm": 0.00015077627904247493,
+      "learning_rate": 1.6429249762583096e-06,
+      "loss": 0.0,
+      "step": 19940
+    },
+    {
+      "epoch": 85.00256410256411,
+      "grad_norm": 0.000555673090275377,
+      "learning_rate": 1.6381766381766383e-06,
+      "loss": 0.0,
+      "step": 19950
+    },
+    {
+      "epoch": 85.00299145299145,
+      "grad_norm": 0.0001231930946232751,
+      "learning_rate": 1.6334283000949669e-06,
+      "loss": 0.0,
+      "step": 19960
+    },
+    {
+      "epoch": 85.00341880341881,
+      "grad_norm": 0.0003615697205532342,
+      "learning_rate": 1.6286799620132954e-06,
+      "loss": 0.0,
+      "step": 19970
+    },
+    {
+      "epoch": 85.00384615384615,
+      "grad_norm": 0.00024398553068749607,
+      "learning_rate": 1.623931623931624e-06,
+      "loss": 0.0,
+      "step": 19980
+    },
+    {
+      "epoch": 85.0042735042735,
+      "grad_norm": 0.0009278811048716307,
+      "learning_rate": 1.6191832858499526e-06,
+      "loss": 0.0,
+      "step": 19990
+    },
+    {
+      "epoch": 85.00470085470086,
+      "grad_norm": 0.0006217070040293038,
+      "learning_rate": 1.6144349477682813e-06,
+      "loss": 0.0,
+      "step": 20000
+    },
+    {
+      "epoch": 85.0051282051282,
+      "grad_norm": 0.000226405740249902,
+      "learning_rate": 1.6096866096866098e-06,
+      "loss": 0.0,
+      "step": 20010
+    },
+    {
+      "epoch": 85.00555555555556,
+      "grad_norm": 0.00026999341207556427,
+      "learning_rate": 1.6049382716049383e-06,
+      "loss": 0.0021,
+      "step": 20020
+    },
+    {
+      "epoch": 85.0059829059829,
+      "grad_norm": 0.00012980998144485056,
+      "learning_rate": 1.600189933523267e-06,
+      "loss": 0.0,
+      "step": 20030
+    },
+    {
+      "epoch": 85.00641025641026,
+      "grad_norm": 0.0006714918417856097,
+      "learning_rate": 1.5954415954415956e-06,
+      "loss": 0.0,
+      "step": 20040
+    },
+    {
+      "epoch": 85.0068376068376,
+      "grad_norm": 0.0013149407459422946,
+      "learning_rate": 1.590693257359924e-06,
+      "loss": 0.0,
+      "step": 20050
+    },
+    {
+      "epoch": 85.00726495726495,
+      "grad_norm": 0.00016666753799654543,
+      "learning_rate": 1.5859449192782528e-06,
+      "loss": 0.0,
+      "step": 20060
+    },
+    {
+      "epoch": 85.00769230769231,
+      "grad_norm": 0.0008010066230781376,
+      "learning_rate": 1.5811965811965813e-06,
+      "loss": 0.0557,
+      "step": 20070
+    },
+    {
+      "epoch": 85.00811965811965,
+      "grad_norm": 0.0002666155050974339,
+      "learning_rate": 1.57644824311491e-06,
+      "loss": 0.0,
+      "step": 20080
+    },
+    {
+      "epoch": 85.00854700854701,
+      "grad_norm": 0.00023051415337249637,
+      "learning_rate": 1.5716999050332385e-06,
+      "loss": 0.0,
+      "step": 20090
+    },
+    {
+      "epoch": 85.00897435897436,
+      "grad_norm": 0.0011181911686435342,
+      "learning_rate": 1.566951566951567e-06,
+      "loss": 0.0,
+      "step": 20100
+    },
+    {
+      "epoch": 85.00940170940171,
+      "grad_norm": 0.0008536745444871485,
+      "learning_rate": 1.5622032288698958e-06,
+      "loss": 0.1919,
+      "step": 20110
+    },
+    {
+      "epoch": 85.00982905982906,
+      "grad_norm": 0.0005208790535107255,
+      "learning_rate": 1.5574548907882243e-06,
+      "loss": 0.0,
+      "step": 20120
+    },
+    {
+      "epoch": 85.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.55494499206543,
+      "eval_runtime": 33.6803,
+      "eval_samples_per_second": 0.742,
+      "eval_steps_per_second": 0.742,
+      "step": 20124
+    },
+    {
+      "epoch": 86.00025641025641,
+      "grad_norm": 0.00014588376507163048,
+      "learning_rate": 1.552706552706553e-06,
+      "loss": 0.0,
+      "step": 20130
+    },
+    {
+      "epoch": 86.00068376068376,
+      "grad_norm": 0.00038126404979266226,
+      "learning_rate": 1.5479582146248815e-06,
+      "loss": 0.0,
+      "step": 20140
+    },
+    {
+      "epoch": 86.00111111111111,
+      "grad_norm": 0.0008430579327978194,
+      "learning_rate": 1.54320987654321e-06,
+      "loss": 0.0,
+      "step": 20150
+    },
+    {
+      "epoch": 86.00153846153846,
+      "grad_norm": 0.0002688311506062746,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 0.0,
+      "step": 20160
+    },
+    {
+      "epoch": 86.00196581196582,
+      "grad_norm": 0.00030692145810462534,
+      "learning_rate": 1.5337132003798672e-06,
+      "loss": 0.0,
+      "step": 20170
+    },
+    {
+      "epoch": 86.00239316239316,
+      "grad_norm": 0.00014298380119726062,
+      "learning_rate": 1.5289648622981957e-06,
+      "loss": 0.0,
+      "step": 20180
+    },
+    {
+      "epoch": 86.0028205128205,
+      "grad_norm": 0.0001439500047126785,
+      "learning_rate": 1.5242165242165245e-06,
+      "loss": 0.0,
+      "step": 20190
+    },
+    {
+      "epoch": 86.00324786324786,
+      "grad_norm": 0.0001333106920355931,
+      "learning_rate": 1.519468186134853e-06,
+      "loss": 0.0,
+      "step": 20200
+    },
+    {
+      "epoch": 86.00367521367521,
+      "grad_norm": 0.0004077716148458421,
+      "learning_rate": 1.5147198480531817e-06,
+      "loss": 0.0,
+      "step": 20210
+    },
+    {
+      "epoch": 86.00410256410257,
+      "grad_norm": 0.0005425411509349942,
+      "learning_rate": 1.5099715099715102e-06,
+      "loss": 0.0081,
+      "step": 20220
+    },
+    {
+      "epoch": 86.00452991452991,
+      "grad_norm": 0.0001421132474206388,
+      "learning_rate": 1.5052231718898385e-06,
+      "loss": 0.0,
+      "step": 20230
+    },
+    {
+      "epoch": 86.00495726495727,
+      "grad_norm": 0.0005388124845921993,
+      "learning_rate": 1.5004748338081674e-06,
+      "loss": 0.0002,
+      "step": 20240
+    },
+    {
+      "epoch": 86.00538461538461,
+      "grad_norm": 0.0001884761149995029,
+      "learning_rate": 1.4957264957264957e-06,
+      "loss": 0.0,
+      "step": 20250
+    },
+    {
+      "epoch": 86.00581196581197,
+      "grad_norm": 0.0015297207282856107,
+      "learning_rate": 1.4909781576448246e-06,
+      "loss": 0.0,
+      "step": 20260
+    },
+    {
+      "epoch": 86.00623931623932,
+      "grad_norm": 0.0007038248004391789,
+      "learning_rate": 1.486229819563153e-06,
+      "loss": 0.0,
+      "step": 20270
+    },
+    {
+      "epoch": 86.00666666666666,
+      "grad_norm": 0.00011914034985238686,
+      "learning_rate": 1.4814814814814815e-06,
+      "loss": 0.0,
+      "step": 20280
+    },
+    {
+      "epoch": 86.00709401709402,
+      "grad_norm": 0.0005224226042628288,
+      "learning_rate": 1.4767331433998102e-06,
+      "loss": 0.0,
+      "step": 20290
+    },
+    {
+      "epoch": 86.00752136752136,
+      "grad_norm": 0.00024032553483266383,
+      "learning_rate": 1.4719848053181387e-06,
+      "loss": 0.0,
+      "step": 20300
+    },
+    {
+      "epoch": 86.00794871794872,
+      "grad_norm": 0.0015607475070282817,
+      "learning_rate": 1.4672364672364672e-06,
+      "loss": 0.0,
+      "step": 20310
+    },
+    {
+      "epoch": 86.00837606837607,
+      "grad_norm": 0.000935015850700438,
+      "learning_rate": 1.462488129154796e-06,
+      "loss": 0.0,
+      "step": 20320
+    },
+    {
+      "epoch": 86.00880341880342,
+      "grad_norm": 0.0009676801273599267,
+      "learning_rate": 1.4577397910731244e-06,
+      "loss": 0.0,
+      "step": 20330
+    },
+    {
+      "epoch": 86.00923076923077,
+      "grad_norm": 0.0018314124317839742,
+      "learning_rate": 1.4529914529914531e-06,
+      "loss": 0.0,
+      "step": 20340
+    },
+    {
+      "epoch": 86.00965811965811,
+      "grad_norm": 0.00039655648288317025,
+      "learning_rate": 1.4482431149097816e-06,
+      "loss": 0.0,
+      "step": 20350
+    },
+    {
+      "epoch": 86.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.722276210784912,
+      "eval_runtime": 34.4477,
+      "eval_samples_per_second": 0.726,
+      "eval_steps_per_second": 0.726,
+      "step": 20358
+    },
+    {
+      "epoch": 87.00008547008547,
+      "grad_norm": 0.00019830497330985963,
+      "learning_rate": 1.4434947768281102e-06,
+      "loss": 0.0,
+      "step": 20360
+    },
+    {
+      "epoch": 87.00051282051282,
+      "grad_norm": 0.00021265774557832628,
+      "learning_rate": 1.4387464387464389e-06,
+      "loss": 0.0,
+      "step": 20370
+    },
+    {
+      "epoch": 87.00094017094017,
+      "grad_norm": 0.00013740244321525097,
+      "learning_rate": 1.4339981006647674e-06,
+      "loss": 0.0,
+      "step": 20380
+    },
+    {
+      "epoch": 87.00136752136753,
+      "grad_norm": 0.0008310623816214502,
+      "learning_rate": 1.429249762583096e-06,
+      "loss": 0.0,
+      "step": 20390
+    },
+    {
+      "epoch": 87.00179487179487,
+      "grad_norm": 0.00020110943296458572,
+      "learning_rate": 1.4245014245014246e-06,
+      "loss": 0.0,
+      "step": 20400
+    },
+    {
+      "epoch": 87.00222222222222,
+      "grad_norm": 0.0004781411844305694,
+      "learning_rate": 1.4197530864197531e-06,
+      "loss": 0.0,
+      "step": 20410
+    },
+    {
+      "epoch": 87.00264957264957,
+      "grad_norm": 0.0001626553712412715,
+      "learning_rate": 1.4150047483380818e-06,
+      "loss": 0.0,
+      "step": 20420
+    },
+    {
+      "epoch": 87.00307692307692,
+      "grad_norm": 0.00045715321903117,
+      "learning_rate": 1.4102564102564104e-06,
+      "loss": 0.0,
+      "step": 20430
+    },
+    {
+      "epoch": 87.00350427350428,
+      "grad_norm": 0.00021040246065240353,
+      "learning_rate": 1.4055080721747389e-06,
+      "loss": 0.0,
+      "step": 20440
+    },
+    {
+      "epoch": 87.00393162393162,
+      "grad_norm": 0.00011547945905476809,
+      "learning_rate": 1.4007597340930676e-06,
+      "loss": 0.0,
+      "step": 20450
+    },
+    {
+      "epoch": 87.00435897435898,
+      "grad_norm": 0.00020624134049285203,
+      "learning_rate": 1.396011396011396e-06,
+      "loss": 0.0,
+      "step": 20460
+    },
+    {
+      "epoch": 87.00478632478632,
+      "grad_norm": 0.0008065582369454205,
+      "learning_rate": 1.3912630579297248e-06,
+      "loss": 0.0,
+      "step": 20470
+    },
+    {
+      "epoch": 87.00521367521368,
+      "grad_norm": 0.00020026916172355413,
+      "learning_rate": 1.3865147198480533e-06,
+      "loss": 0.0,
+      "step": 20480
+    },
+    {
+      "epoch": 87.00564102564103,
+      "grad_norm": 0.00014089647447690368,
+      "learning_rate": 1.3817663817663818e-06,
+      "loss": 0.0,
+      "step": 20490
+    },
+    {
+      "epoch": 87.00606837606837,
+      "grad_norm": 0.0007202147389762104,
+      "learning_rate": 1.3770180436847105e-06,
+      "loss": 0.0,
+      "step": 20500
+    },
+    {
+      "epoch": 87.00649572649573,
+      "grad_norm": 0.0001863235083874315,
+      "learning_rate": 1.372269705603039e-06,
+      "loss": 0.0,
+      "step": 20510
+    },
+    {
+      "epoch": 87.00692307692307,
+      "grad_norm": 0.00013089961430523545,
+      "learning_rate": 1.3675213675213678e-06,
+      "loss": 0.0,
+      "step": 20520
+    },
+    {
+      "epoch": 87.00735042735043,
+      "grad_norm": 0.000519382010679692,
+      "learning_rate": 1.3627730294396963e-06,
+      "loss": 0.0,
+      "step": 20530
+    },
+    {
+      "epoch": 87.00777777777778,
+      "grad_norm": 0.00013078245683573186,
+      "learning_rate": 1.3580246913580248e-06,
+      "loss": 0.0,
+      "step": 20540
+    },
+    {
+      "epoch": 87.00820512820513,
+      "grad_norm": 0.0006038915598765016,
+      "learning_rate": 1.3532763532763535e-06,
+      "loss": 0.0,
+      "step": 20550
+    },
+    {
+      "epoch": 87.00863247863248,
+      "grad_norm": 0.0002612897951621562,
+      "learning_rate": 1.348528015194682e-06,
+      "loss": 0.0,
+      "step": 20560
+    },
+    {
+      "epoch": 87.00905982905982,
+      "grad_norm": 0.00011701687617460266,
+      "learning_rate": 1.3437796771130105e-06,
+      "loss": 0.3334,
+      "step": 20570
+    },
+    {
+      "epoch": 87.00948717948718,
+      "grad_norm": 0.0023586824536323547,
+      "learning_rate": 1.3390313390313392e-06,
+      "loss": 0.0,
+      "step": 20580
+    },
+    {
+      "epoch": 87.00991452991452,
+      "grad_norm": 0.00032859586644917727,
+      "learning_rate": 1.3342830009496678e-06,
+      "loss": 0.0,
+      "step": 20590
+    },
+    {
+      "epoch": 87.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.909550189971924,
+      "eval_runtime": 34.96,
+      "eval_samples_per_second": 0.715,
+      "eval_steps_per_second": 0.715,
+      "step": 20592
+    },
+    {
+      "epoch": 88.00034188034188,
+      "grad_norm": 0.00037892567343078554,
+      "learning_rate": 1.3295346628679965e-06,
+      "loss": 0.0,
+      "step": 20600
+    },
+    {
+      "epoch": 88.00076923076924,
+      "grad_norm": 0.00016175136261153966,
+      "learning_rate": 1.324786324786325e-06,
+      "loss": 0.0,
+      "step": 20610
+    },
+    {
+      "epoch": 88.00119658119658,
+      "grad_norm": 0.00011514942161738873,
+      "learning_rate": 1.3200379867046533e-06,
+      "loss": 0.0,
+      "step": 20620
+    },
+    {
+      "epoch": 88.00162393162393,
+      "grad_norm": 0.000466162251541391,
+      "learning_rate": 1.3152896486229822e-06,
+      "loss": 0.0,
+      "step": 20630
+    },
+    {
+      "epoch": 88.00205128205128,
+      "grad_norm": 0.0005339948693290353,
+      "learning_rate": 1.3105413105413107e-06,
+      "loss": 0.0,
+      "step": 20640
+    },
+    {
+      "epoch": 88.00247863247863,
+      "grad_norm": 0.00011431486200308427,
+      "learning_rate": 1.305792972459639e-06,
+      "loss": 0.0,
+      "step": 20650
+    },
+    {
+      "epoch": 88.00290598290599,
+      "grad_norm": 0.00010979742364725098,
+      "learning_rate": 1.301044634377968e-06,
+      "loss": 0.0,
+      "step": 20660
+    },
+    {
+      "epoch": 88.00333333333333,
+      "grad_norm": 0.001283150864765048,
+      "learning_rate": 1.2962962962962962e-06,
+      "loss": 0.0,
+      "step": 20670
+    },
+    {
+      "epoch": 88.00376068376069,
+      "grad_norm": 0.0002451012551318854,
+      "learning_rate": 1.2915479582146252e-06,
+      "loss": 0.0,
+      "step": 20680
+    },
+    {
+      "epoch": 88.00418803418803,
+      "grad_norm": 0.00014794590242672712,
+      "learning_rate": 1.2867996201329535e-06,
+      "loss": 0.0,
+      "step": 20690
+    },
+    {
+      "epoch": 88.00461538461539,
+      "grad_norm": 0.0005672202096320689,
+      "learning_rate": 1.282051282051282e-06,
+      "loss": 0.0,
+      "step": 20700
+    },
+    {
+      "epoch": 88.00504273504274,
+      "grad_norm": 0.00013605415006168187,
+      "learning_rate": 1.2773029439696107e-06,
+      "loss": 0.0,
+      "step": 20710
+    },
+    {
+      "epoch": 88.00547008547008,
+      "grad_norm": 0.00011451664613559842,
+      "learning_rate": 1.2725546058879392e-06,
+      "loss": 0.0,
+      "step": 20720
+    },
+    {
+      "epoch": 88.00589743589744,
+      "grad_norm": 0.00019140506628900766,
+      "learning_rate": 1.267806267806268e-06,
+      "loss": 0.0,
+      "step": 20730
+    },
+    {
+      "epoch": 88.00632478632478,
+      "grad_norm": 0.00013691678759641945,
+      "learning_rate": 1.2630579297245964e-06,
+      "loss": 0.0,
+      "step": 20740
+    },
+    {
+      "epoch": 88.00675213675214,
+      "grad_norm": 0.00016725384921301156,
+      "learning_rate": 1.258309591642925e-06,
+      "loss": 0.0,
+      "step": 20750
+    },
+    {
+      "epoch": 88.00717948717949,
+      "grad_norm": 0.00016475172014907002,
+      "learning_rate": 1.2535612535612537e-06,
+      "loss": 0.0,
+      "step": 20760
+    },
+    {
+      "epoch": 88.00760683760684,
+      "grad_norm": 0.00015170658298302442,
+      "learning_rate": 1.2488129154795822e-06,
+      "loss": 0.0,
+      "step": 20770
+    },
+    {
+      "epoch": 88.00803418803419,
+      "grad_norm": 0.012962790206074715,
+      "learning_rate": 1.244064577397911e-06,
+      "loss": 0.0,
+      "step": 20780
+    },
+    {
+      "epoch": 88.00846153846153,
+      "grad_norm": 0.00019268876349087805,
+      "learning_rate": 1.2393162393162394e-06,
+      "loss": 0.0,
+      "step": 20790
+    },
+    {
+      "epoch": 88.00888888888889,
+      "grad_norm": 0.00012487791536841542,
+      "learning_rate": 1.234567901234568e-06,
+      "loss": 0.0,
+      "step": 20800
+    },
+    {
+      "epoch": 88.00931623931623,
+      "grad_norm": 0.009372721426188946,
+      "learning_rate": 1.2298195631528966e-06,
+      "loss": 0.0,
+      "step": 20810
+    },
+    {
+      "epoch": 88.0097435897436,
+      "grad_norm": 0.0001522890670457855,
+      "learning_rate": 1.2250712250712251e-06,
+      "loss": 0.0,
+      "step": 20820
+    },
+    {
+      "epoch": 88.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.991762638092041,
+      "eval_runtime": 34.8658,
+      "eval_samples_per_second": 0.717,
+      "eval_steps_per_second": 0.717,
+      "step": 20826
+    },
+    {
+      "epoch": 89.00017094017095,
+      "grad_norm": 1.7047139406204224,
+      "learning_rate": 1.2203228869895539e-06,
+      "loss": 0.0003,
+      "step": 20830
+    },
+    {
+      "epoch": 89.00059829059829,
+      "grad_norm": 0.004476090893149376,
+      "learning_rate": 1.2155745489078824e-06,
+      "loss": 0.0,
+      "step": 20840
+    },
+    {
+      "epoch": 89.00102564102563,
+      "grad_norm": 0.0009234125609509647,
+      "learning_rate": 1.2108262108262109e-06,
+      "loss": 0.0,
+      "step": 20850
+    },
+    {
+      "epoch": 89.001452991453,
+      "grad_norm": 0.0005792377050966024,
+      "learning_rate": 1.2060778727445396e-06,
+      "loss": 0.0,
+      "step": 20860
+    },
+    {
+      "epoch": 89.00188034188034,
+      "grad_norm": 0.0001534123730380088,
+      "learning_rate": 1.2013295346628681e-06,
+      "loss": 0.0,
+      "step": 20870
+    },
+    {
+      "epoch": 89.0023076923077,
+      "grad_norm": 0.0013493632432073355,
+      "learning_rate": 1.1965811965811968e-06,
+      "loss": 0.0,
+      "step": 20880
+    },
+    {
+      "epoch": 89.00273504273504,
+      "grad_norm": 0.00017192041559610516,
+      "learning_rate": 1.1918328584995251e-06,
+      "loss": 0.0,
+      "step": 20890
+    },
+    {
+      "epoch": 89.0031623931624,
+      "grad_norm": 0.00012882999726571143,
+      "learning_rate": 1.1870845204178538e-06,
+      "loss": 0.0,
+      "step": 20900
+    },
+    {
+      "epoch": 89.00358974358974,
+      "grad_norm": 0.0013170551974326372,
+      "learning_rate": 1.1823361823361824e-06,
+      "loss": 0.0,
+      "step": 20910
+    },
+    {
+      "epoch": 89.0040170940171,
+      "grad_norm": 0.0004404323117341846,
+      "learning_rate": 1.177587844254511e-06,
+      "loss": 0.0,
+      "step": 20920
+    },
+    {
+      "epoch": 89.00444444444445,
+      "grad_norm": 0.00011723486386472359,
+      "learning_rate": 1.1728395061728396e-06,
+      "loss": 0.0015,
+      "step": 20930
+    },
+    {
+      "epoch": 89.00487179487179,
+      "grad_norm": 0.00012327823787927628,
+      "learning_rate": 1.168091168091168e-06,
+      "loss": 0.0,
+      "step": 20940
+    },
+    {
+      "epoch": 89.00529914529915,
+      "grad_norm": 0.00013142797979526222,
+      "learning_rate": 1.1633428300094968e-06,
+      "loss": 0.0,
+      "step": 20950
+    },
+    {
+      "epoch": 89.00572649572649,
+      "grad_norm": 0.0001497309422120452,
+      "learning_rate": 1.1585944919278253e-06,
+      "loss": 0.0,
+      "step": 20960
+    },
+    {
+      "epoch": 89.00615384615385,
+      "grad_norm": 0.0004792478575836867,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 0.0,
+      "step": 20970
+    },
+    {
+      "epoch": 89.0065811965812,
+      "grad_norm": 0.0001388894597766921,
+      "learning_rate": 1.1490978157644825e-06,
+      "loss": 0.0,
+      "step": 20980
+    },
+    {
+      "epoch": 89.00700854700855,
+      "grad_norm": 0.00023399751808028668,
+      "learning_rate": 1.144349477682811e-06,
+      "loss": 0.0,
+      "step": 20990
+    },
+    {
+      "epoch": 89.0074358974359,
+      "grad_norm": 0.00014475570060312748,
+      "learning_rate": 1.1396011396011398e-06,
+      "loss": 0.0,
+      "step": 21000
+    },
+    {
+      "epoch": 89.00786324786324,
+      "grad_norm": 0.00012018286361126229,
+      "learning_rate": 1.1348528015194683e-06,
+      "loss": 0.0,
+      "step": 21010
+    },
+    {
+      "epoch": 89.0082905982906,
+      "grad_norm": 0.00017406410188414156,
+      "learning_rate": 1.1301044634377968e-06,
+      "loss": 0.0002,
+      "step": 21020
+    },
+    {
+      "epoch": 89.00871794871794,
+      "grad_norm": 0.0001829592656577006,
+      "learning_rate": 1.1253561253561255e-06,
+      "loss": 0.0,
+      "step": 21030
+    },
+    {
+      "epoch": 89.0091452991453,
+      "grad_norm": 0.0005391041049733758,
+      "learning_rate": 1.120607787274454e-06,
+      "loss": 0.0,
+      "step": 21040
+    },
+    {
+      "epoch": 89.00957264957265,
+      "grad_norm": 0.00044315739069133997,
+      "learning_rate": 1.1158594491927827e-06,
+      "loss": 0.0,
+      "step": 21050
+    },
+    {
+      "epoch": 89.01,
+      "grad_norm": 0.00018860511772800237,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 0.0,
+      "step": 21060
+    },
+    {
+      "epoch": 89.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 7.22465705871582,
+      "eval_runtime": 35.0687,
+      "eval_samples_per_second": 0.713,
+      "eval_steps_per_second": 0.713,
+      "step": 21060
+    },
+    {
+      "epoch": 90.00042735042734,
+      "grad_norm": 0.00021203322103247046,
+      "learning_rate": 1.1063627730294398e-06,
+      "loss": 0.0,
+      "step": 21070
+    },
+    {
+      "epoch": 90.0008547008547,
+      "grad_norm": 0.00013690422929357737,
+      "learning_rate": 1.1016144349477685e-06,
+      "loss": 0.0,
+      "step": 21080
+    },
+    {
+      "epoch": 90.00128205128205,
+      "grad_norm": 0.0015827484894543886,
+      "learning_rate": 1.096866096866097e-06,
+      "loss": 0.6422,
+      "step": 21090
+    },
+    {
+      "epoch": 90.0017094017094,
+      "grad_norm": 0.00015395709488075227,
+      "learning_rate": 1.0921177587844257e-06,
+      "loss": 0.0,
+      "step": 21100
+    },
+    {
+      "epoch": 90.00213675213675,
+      "grad_norm": 0.0011581846047192812,
+      "learning_rate": 1.087369420702754e-06,
+      "loss": 0.0,
+      "step": 21110
+    },
+    {
+      "epoch": 90.00256410256411,
+      "grad_norm": 0.00010031823330791667,
+      "learning_rate": 1.0826210826210827e-06,
+      "loss": 0.0,
+      "step": 21120
+    },
+    {
+      "epoch": 90.00299145299145,
+      "grad_norm": 0.00012422981671988964,
+      "learning_rate": 1.0778727445394112e-06,
+      "loss": 0.0,
+      "step": 21130
+    },
+    {
+      "epoch": 90.00341880341881,
+      "grad_norm": 0.009594716131687164,
+      "learning_rate": 1.07312440645774e-06,
+      "loss": 0.0,
+      "step": 21140
+    },
+    {
+      "epoch": 90.00384615384615,
+      "grad_norm": 0.010623271577060223,
+      "learning_rate": 1.0683760683760685e-06,
+      "loss": 0.0,
+      "step": 21150
+    },
+    {
+      "epoch": 90.0042735042735,
+      "grad_norm": 0.0004154906782787293,
+      "learning_rate": 1.063627730294397e-06,
+      "loss": 0.0,
+      "step": 21160
+    },
+    {
+      "epoch": 90.00470085470086,
+      "grad_norm": 0.0005235851858742535,
+      "learning_rate": 1.0588793922127257e-06,
+      "loss": 0.0,
+      "step": 21170
+    },
+    {
+      "epoch": 90.0051282051282,
+      "grad_norm": 0.0006127028027549386,
+      "learning_rate": 1.0541310541310542e-06,
+      "loss": 0.0,
+      "step": 21180
+    },
+    {
+      "epoch": 90.00555555555556,
+      "grad_norm": 0.00012018286361126229,
+      "learning_rate": 1.0493827160493827e-06,
+      "loss": 0.0,
+      "step": 21190
+    },
+    {
+      "epoch": 90.0059829059829,
+      "grad_norm": 0.00017059137462638319,
+      "learning_rate": 1.0446343779677114e-06,
+      "loss": 0.0,
+      "step": 21200
+    },
+    {
+      "epoch": 90.00641025641026,
+      "grad_norm": 0.00011609737703111023,
+      "learning_rate": 1.03988603988604e-06,
+      "loss": 0.0012,
+      "step": 21210
+    },
+    {
+      "epoch": 90.0068376068376,
+      "grad_norm": 0.00016542985395062715,
+      "learning_rate": 1.0351377018043687e-06,
+      "loss": 0.0001,
+      "step": 21220
+    },
+    {
+      "epoch": 90.00726495726495,
+      "grad_norm": 0.22980256378650665,
+      "learning_rate": 1.0303893637226972e-06,
+      "loss": 0.0001,
+      "step": 21230
+    },
+    {
+      "epoch": 90.00769230769231,
+      "grad_norm": 0.0003457261191215366,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 0.0,
+      "step": 21240
+    },
+    {
+      "epoch": 90.00811965811965,
+      "grad_norm": 0.0006093020201660693,
+      "learning_rate": 1.0208926875593544e-06,
+      "loss": 0.0,
+      "step": 21250
+    },
+    {
+      "epoch": 90.00854700854701,
+      "grad_norm": 0.00011088740575360134,
+      "learning_rate": 1.016144349477683e-06,
+      "loss": 0.0,
+      "step": 21260
+    },
+    {
+      "epoch": 90.00897435897436,
+      "grad_norm": 0.0004857526218984276,
+      "learning_rate": 1.0113960113960116e-06,
+      "loss": 0.0,
+      "step": 21270
+    },
+    {
+      "epoch": 90.00940170940171,
+      "grad_norm": 0.00013570766896009445,
+      "learning_rate": 1.00664767331434e-06,
+      "loss": 0.0,
+      "step": 21280
+    },
+    {
+      "epoch": 90.00982905982906,
+      "grad_norm": 0.0014373520389199257,
+      "learning_rate": 1.0018993352326686e-06,
+      "loss": 0.0001,
+      "step": 21290
+    },
+    {
+      "epoch": 90.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 7.226686954498291,
+      "eval_runtime": 37.4838,
+      "eval_samples_per_second": 0.667,
+      "eval_steps_per_second": 0.667,
+      "step": 21294
+    },
+    {
+      "epoch": 91.00025641025641,
+      "grad_norm": 0.00011334038572385907,
+      "learning_rate": 9.971509971509974e-07,
+      "loss": 0.0,
+      "step": 21300
+    },
+    {
+      "epoch": 91.00068376068376,
+      "grad_norm": 0.00015362763951998204,
+      "learning_rate": 9.924026590693259e-07,
+      "loss": 0.0,
+      "step": 21310
+    },
+    {
+      "epoch": 91.00111111111111,
+      "grad_norm": 0.0003192702424712479,
+      "learning_rate": 9.876543209876544e-07,
+      "loss": 0.0,
+      "step": 21320
+    },
+    {
+      "epoch": 91.00153846153846,
+      "grad_norm": 0.0005757631151936948,
+      "learning_rate": 9.829059829059829e-07,
+      "loss": 0.0,
+      "step": 21330
+    },
+    {
+      "epoch": 91.00196581196582,
+      "grad_norm": 0.0004981070524081588,
+      "learning_rate": 9.781576448243116e-07,
+      "loss": 0.0,
+      "step": 21340
+    },
+    {
+      "epoch": 91.00239316239316,
+      "grad_norm": 0.000131376669742167,
+      "learning_rate": 9.734093067426401e-07,
+      "loss": 0.0,
+      "step": 21350
+    },
+    {
+      "epoch": 91.0028205128205,
+      "grad_norm": 1261.06396484375,
+      "learning_rate": 9.686609686609686e-07,
+      "loss": 0.3664,
+      "step": 21360
+    },
+    {
+      "epoch": 91.00324786324786,
+      "grad_norm": 0.00013327557826414704,
+      "learning_rate": 9.639126305792973e-07,
+      "loss": 0.0,
+      "step": 21370
+    },
+    {
+      "epoch": 91.00367521367521,
+      "grad_norm": 0.0001816957665141672,
+      "learning_rate": 9.591642924976258e-07,
+      "loss": 0.0,
+      "step": 21380
+    },
+    {
+      "epoch": 91.00410256410257,
+      "grad_norm": 0.004342993255704641,
+      "learning_rate": 9.544159544159546e-07,
+      "loss": 0.0,
+      "step": 21390
+    },
+    {
+      "epoch": 91.00452991452991,
+      "grad_norm": 0.0002144295140169561,
+      "learning_rate": 9.496676163342832e-07,
+      "loss": 0.0,
+      "step": 21400
+    },
+    {
+      "epoch": 91.00495726495727,
+      "grad_norm": 0.0005280747427605093,
+      "learning_rate": 9.449192782526116e-07,
+      "loss": 0.0001,
+      "step": 21410
+    },
+    {
+      "epoch": 91.00538461538461,
+      "grad_norm": 0.0002650852547958493,
+      "learning_rate": 9.401709401709402e-07,
+      "loss": 0.0,
+      "step": 21420
+    },
+    {
+      "epoch": 91.00581196581197,
+      "grad_norm": 0.0003367229946888983,
+      "learning_rate": 9.354226020892688e-07,
+      "loss": 0.0,
+      "step": 21430
+    },
+    {
+      "epoch": 91.00623931623932,
+      "grad_norm": 0.00010731956717791036,
+      "learning_rate": 9.306742640075974e-07,
+      "loss": 0.0,
+      "step": 21440
+    },
+    {
+      "epoch": 91.00666666666666,
+      "grad_norm": 0.00010221748379990458,
+      "learning_rate": 9.259259259259259e-07,
+      "loss": 0.0,
+      "step": 21450
+    },
+    {
+      "epoch": 91.00709401709402,
+      "grad_norm": 0.00013446349475998431,
+      "learning_rate": 9.211775878442545e-07,
+      "loss": 0.0,
+      "step": 21460
+    },
+    {
+      "epoch": 91.00752136752136,
+      "grad_norm": 0.00018187399837188423,
+      "learning_rate": 9.164292497625832e-07,
+      "loss": 0.0,
+      "step": 21470
+    },
+    {
+      "epoch": 91.00794871794872,
+      "grad_norm": 0.00015760202950332314,
+      "learning_rate": 9.116809116809118e-07,
+      "loss": 0.0,
+      "step": 21480
+    },
+    {
+      "epoch": 91.00837606837607,
+      "grad_norm": 0.00026204713503830135,
+      "learning_rate": 9.069325735992403e-07,
+      "loss": 0.0,
+      "step": 21490
+    },
+    {
+      "epoch": 91.00880341880342,
+      "grad_norm": 0.00038499056245200336,
+      "learning_rate": 9.021842355175689e-07,
+      "loss": 0.0,
+      "step": 21500
+    },
+    {
+      "epoch": 91.00923076923077,
+      "grad_norm": 0.000741164549253881,
+      "learning_rate": 8.974358974358975e-07,
+      "loss": 0.0,
+      "step": 21510
+    },
+    {
+      "epoch": 91.00965811965811,
+      "grad_norm": 0.00012043473543599248,
+      "learning_rate": 8.926875593542261e-07,
+      "loss": 0.0,
+      "step": 21520
+    },
+    {
+      "epoch": 91.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.982587814331055,
+      "eval_runtime": 37.2182,
+      "eval_samples_per_second": 0.672,
+      "eval_steps_per_second": 0.672,
+      "step": 21528
+    },
+    {
+      "epoch": 92.00008547008547,
+      "grad_norm": 0.0001411245611961931,
+      "learning_rate": 8.879392212725547e-07,
+      "loss": 0.0,
+      "step": 21530
+    },
+    {
+      "epoch": 92.00051282051282,
+      "grad_norm": 0.0017277189763262868,
+      "learning_rate": 8.831908831908833e-07,
+      "loss": 0.0,
+      "step": 21540
+    },
+    {
+      "epoch": 92.00094017094017,
+      "grad_norm": 0.00011866500426549464,
+      "learning_rate": 8.784425451092119e-07,
+      "loss": 0.0,
+      "step": 21550
+    },
+    {
+      "epoch": 92.00136752136753,
+      "grad_norm": 0.00031862963805906475,
+      "learning_rate": 8.736942070275405e-07,
+      "loss": 0.0,
+      "step": 21560
+    },
+    {
+      "epoch": 92.00179487179487,
+      "grad_norm": 0.00011709554382832721,
+      "learning_rate": 8.689458689458691e-07,
+      "loss": 0.0,
+      "step": 21570
+    },
+    {
+      "epoch": 92.00222222222222,
+      "grad_norm": 0.0005032451008446515,
+      "learning_rate": 8.641975308641976e-07,
+      "loss": 0.0,
+      "step": 21580
+    },
+    {
+      "epoch": 92.00264957264957,
+      "grad_norm": 0.0010287113254889846,
+      "learning_rate": 8.594491927825262e-07,
+      "loss": 0.0,
+      "step": 21590
+    },
+    {
+      "epoch": 92.00307692307692,
+      "grad_norm": 0.0004769675724674016,
+      "learning_rate": 8.547008547008548e-07,
+      "loss": 0.0,
+      "step": 21600
+    },
+    {
+      "epoch": 92.00350427350428,
+      "grad_norm": 0.00032892791205085814,
+      "learning_rate": 8.499525166191834e-07,
+      "loss": 0.0,
+      "step": 21610
+    },
+    {
+      "epoch": 92.00393162393162,
+      "grad_norm": 0.0001310220395680517,
+      "learning_rate": 8.452041785375118e-07,
+      "loss": 0.0,
+      "step": 21620
+    },
+    {
+      "epoch": 92.00435897435898,
+      "grad_norm": 0.0010653804056346416,
+      "learning_rate": 8.404558404558405e-07,
+      "loss": 0.0,
+      "step": 21630
+    },
+    {
+      "epoch": 92.00478632478632,
+      "grad_norm": 0.00020127870084252208,
+      "learning_rate": 8.357075023741691e-07,
+      "loss": 0.0,
+      "step": 21640
+    },
+    {
+      "epoch": 92.00521367521368,
+      "grad_norm": 0.0025258706882596016,
+      "learning_rate": 8.309591642924977e-07,
+      "loss": 0.0,
+      "step": 21650
+    },
+    {
+      "epoch": 92.00564102564103,
+      "grad_norm": 0.00015467533376067877,
+      "learning_rate": 8.262108262108262e-07,
+      "loss": 0.9533,
+      "step": 21660
+    },
+    {
+      "epoch": 92.00606837606837,
+      "grad_norm": 0.00014860654482617974,
+      "learning_rate": 8.214624881291548e-07,
+      "loss": 0.0,
+      "step": 21670
+    },
+    {
+      "epoch": 92.00649572649573,
+      "grad_norm": 0.0005292973946779966,
+      "learning_rate": 8.167141500474834e-07,
+      "loss": 0.0,
+      "step": 21680
+    },
+    {
+      "epoch": 92.00692307692307,
+      "grad_norm": 0.00015186695964075625,
+      "learning_rate": 8.11965811965812e-07,
+      "loss": 0.0,
+      "step": 21690
+    },
+    {
+      "epoch": 92.00735042735043,
+      "grad_norm": 0.00021409436885733157,
+      "learning_rate": 8.072174738841407e-07,
+      "loss": 0.0,
+      "step": 21700
+    },
+    {
+      "epoch": 92.00777777777778,
+      "grad_norm": 0.0007851360714994371,
+      "learning_rate": 8.024691358024692e-07,
+      "loss": 0.0,
+      "step": 21710
+    },
+    {
+      "epoch": 92.00820512820513,
+      "grad_norm": 0.00014987800386734307,
+      "learning_rate": 7.977207977207978e-07,
+      "loss": 0.0,
+      "step": 21720
+    },
+    {
+      "epoch": 92.00863247863248,
+      "grad_norm": 0.00021006350289098918,
+      "learning_rate": 7.929724596391264e-07,
+      "loss": 0.0001,
+      "step": 21730
+    },
+    {
+      "epoch": 92.00905982905982,
+      "grad_norm": 0.00011848520807689056,
+      "learning_rate": 7.88224121557455e-07,
+      "loss": 0.0,
+      "step": 21740
+    },
+    {
+      "epoch": 92.00948717948718,
+      "grad_norm": 0.0001388939272146672,
+      "learning_rate": 7.834757834757835e-07,
+      "loss": 0.0,
+      "step": 21750
+    },
+    {
+      "epoch": 92.00991452991452,
+      "grad_norm": 0.00011886980064446107,
+      "learning_rate": 7.787274453941121e-07,
+      "loss": 0.0,
+      "step": 21760
+    },
+    {
+      "epoch": 92.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.638452529907227,
+      "eval_runtime": 36.9317,
+      "eval_samples_per_second": 0.677,
+      "eval_steps_per_second": 0.677,
+      "step": 21762
+    },
+    {
+      "epoch": 93.00034188034188,
+      "grad_norm": 0.00011863345571327955,
+      "learning_rate": 7.739791073124407e-07,
+      "loss": 0.0,
+      "step": 21770
+    },
+    {
+      "epoch": 93.00076923076924,
+      "grad_norm": 0.000148111954331398,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 0.0,
+      "step": 21780
+    },
+    {
+      "epoch": 93.00119658119658,
+      "grad_norm": 0.0006991415284574032,
+      "learning_rate": 7.644824311490979e-07,
+      "loss": 0.0,
+      "step": 21790
+    },
+    {
+      "epoch": 93.00162393162393,
+      "grad_norm": 0.006693857256323099,
+      "learning_rate": 7.597340930674265e-07,
+      "loss": 0.0,
+      "step": 21800
+    },
+    {
+      "epoch": 93.00205128205128,
+      "grad_norm": 0.0005863769329153001,
+      "learning_rate": 7.549857549857551e-07,
+      "loss": 0.0,
+      "step": 21810
+    },
+    {
+      "epoch": 93.00247863247863,
+      "grad_norm": 0.00011994949454674497,
+      "learning_rate": 7.502374169040837e-07,
+      "loss": 0.0,
+      "step": 21820
+    },
+    {
+      "epoch": 93.00290598290599,
+      "grad_norm": 0.00011668866500258446,
+      "learning_rate": 7.454890788224123e-07,
+      "loss": 0.0001,
+      "step": 21830
+    },
+    {
+      "epoch": 93.00333333333333,
+      "grad_norm": 0.00011160215217387304,
+      "learning_rate": 7.407407407407407e-07,
+      "loss": 0.0,
+      "step": 21840
+    },
+    {
+      "epoch": 93.00376068376069,
+      "grad_norm": 0.0001440770720364526,
+      "learning_rate": 7.359924026590693e-07,
+      "loss": 0.0,
+      "step": 21850
+    },
+    {
+      "epoch": 93.00418803418803,
+      "grad_norm": 0.00012972467811778188,
+      "learning_rate": 7.31244064577398e-07,
+      "loss": 0.0,
+      "step": 21860
+    },
+    {
+      "epoch": 93.00461538461539,
+      "grad_norm": 0.0002495836524758488,
+      "learning_rate": 7.264957264957266e-07,
+      "loss": 0.0,
+      "step": 21870
+    },
+    {
+      "epoch": 93.00504273504274,
+      "grad_norm": 0.00011603141319938004,
+      "learning_rate": 7.217473884140551e-07,
+      "loss": 0.0,
+      "step": 21880
+    },
+    {
+      "epoch": 93.00547008547008,
+      "grad_norm": 0.0005587959312833846,
+      "learning_rate": 7.169990503323837e-07,
+      "loss": 0.0,
+      "step": 21890
+    },
+    {
+      "epoch": 93.00589743589744,
+      "grad_norm": 0.00015568920935038477,
+      "learning_rate": 7.122507122507123e-07,
+      "loss": 0.0,
+      "step": 21900
+    },
+    {
+      "epoch": 93.00632478632478,
+      "grad_norm": 0.00015569105744361877,
+      "learning_rate": 7.075023741690409e-07,
+      "loss": 0.0,
+      "step": 21910
+    },
+    {
+      "epoch": 93.00675213675214,
+      "grad_norm": 0.00011201628512935713,
+      "learning_rate": 7.027540360873694e-07,
+      "loss": 0.0,
+      "step": 21920
+    },
+    {
+      "epoch": 93.00717948717949,
+      "grad_norm": 0.0012152871349826455,
+      "learning_rate": 6.98005698005698e-07,
+      "loss": 0.0,
+      "step": 21930
+    },
+    {
+      "epoch": 93.00760683760684,
+      "grad_norm": 0.00011229109804844484,
+      "learning_rate": 6.932573599240267e-07,
+      "loss": 0.0,
+      "step": 21940
+    },
+    {
+      "epoch": 93.00803418803419,
+      "grad_norm": 0.0010557627538219094,
+      "learning_rate": 6.885090218423553e-07,
+      "loss": 0.0,
+      "step": 21950
+    },
+    {
+      "epoch": 93.00846153846153,
+      "grad_norm": 0.0005080616101622581,
+      "learning_rate": 6.837606837606839e-07,
+      "loss": 0.0,
+      "step": 21960
+    },
+    {
+      "epoch": 93.00888888888889,
+      "grad_norm": 0.00011757034371839836,
+      "learning_rate": 6.790123456790124e-07,
+      "loss": 0.0001,
+      "step": 21970
+    },
+    {
+      "epoch": 93.00931623931623,
+      "grad_norm": 0.00048101996071636677,
+      "learning_rate": 6.74264007597341e-07,
+      "loss": 0.0,
+      "step": 21980
+    },
+    {
+      "epoch": 93.0097435897436,
+      "grad_norm": 0.00011343754886183888,
+      "learning_rate": 6.695156695156696e-07,
+      "loss": 0.792,
+      "step": 21990
+    },
+    {
+      "epoch": 93.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.402031898498535,
+      "eval_runtime": 34.6678,
+      "eval_samples_per_second": 0.721,
+      "eval_steps_per_second": 0.721,
+      "step": 21996
+    },
+    {
+      "epoch": 94.00017094017095,
+      "grad_norm": 0.00010894170554820448,
+      "learning_rate": 6.647673314339982e-07,
+      "loss": 0.0,
+      "step": 22000
+    },
+    {
+      "epoch": 94.00059829059829,
+      "grad_norm": 0.00013790665252599865,
+      "learning_rate": 6.600189933523266e-07,
+      "loss": 0.0,
+      "step": 22010
+    },
+    {
+      "epoch": 94.00102564102563,
+      "grad_norm": 0.00014016970817465335,
+      "learning_rate": 6.552706552706554e-07,
+      "loss": 0.0,
+      "step": 22020
+    },
+    {
+      "epoch": 94.001452991453,
+      "grad_norm": 0.002228393219411373,
+      "learning_rate": 6.50522317188984e-07,
+      "loss": 0.0,
+      "step": 22030
+    },
+    {
+      "epoch": 94.00188034188034,
+      "grad_norm": 0.00015126651851460338,
+      "learning_rate": 6.457739791073126e-07,
+      "loss": 0.0,
+      "step": 22040
+    },
+    {
+      "epoch": 94.0023076923077,
+      "grad_norm": 0.008348544128239155,
+      "learning_rate": 6.41025641025641e-07,
+      "loss": 0.0,
+      "step": 22050
+    },
+    {
+      "epoch": 94.00273504273504,
+      "grad_norm": 0.00011175816325703636,
+      "learning_rate": 6.362773029439696e-07,
+      "loss": 0.0001,
+      "step": 22060
+    },
+    {
+      "epoch": 94.0031623931624,
+      "grad_norm": 0.00012174518633401021,
+      "learning_rate": 6.315289648622982e-07,
+      "loss": 0.0,
+      "step": 22070
+    },
+    {
+      "epoch": 94.00358974358974,
+      "grad_norm": 0.00026770823751576245,
+      "learning_rate": 6.267806267806268e-07,
+      "loss": 0.0,
+      "step": 22080
+    },
+    {
+      "epoch": 94.0040170940171,
+      "grad_norm": 0.00019762790179811418,
+      "learning_rate": 6.220322886989554e-07,
+      "loss": 0.0158,
+      "step": 22090
+    },
+    {
+      "epoch": 94.00444444444445,
+      "grad_norm": 0.00017857948841992766,
+      "learning_rate": 6.17283950617284e-07,
+      "loss": 0.0001,
+      "step": 22100
+    },
+    {
+      "epoch": 94.00487179487179,
+      "grad_norm": 0.00011318692850181833,
+      "learning_rate": 6.125356125356126e-07,
+      "loss": 0.0,
+      "step": 22110
+    },
+    {
+      "epoch": 94.00529914529915,
+      "grad_norm": 1054.87451171875,
+      "learning_rate": 6.077872744539412e-07,
+      "loss": 0.1906,
+      "step": 22120
+    },
+    {
+      "epoch": 94.00572649572649,
+      "grad_norm": 0.0002740290074143559,
+      "learning_rate": 6.030389363722698e-07,
+      "loss": 0.0,
+      "step": 22130
+    },
+    {
+      "epoch": 94.00615384615385,
+      "grad_norm": 0.00017096682859119028,
+      "learning_rate": 5.982905982905984e-07,
+      "loss": 0.0,
+      "step": 22140
+    },
+    {
+      "epoch": 94.0065811965812,
+      "grad_norm": 0.00021798326633870602,
+      "learning_rate": 5.935422602089269e-07,
+      "loss": 0.0,
+      "step": 22150
+    },
+    {
+      "epoch": 94.00700854700855,
+      "grad_norm": 0.00105681037530303,
+      "learning_rate": 5.887939221272555e-07,
+      "loss": 0.0,
+      "step": 22160
+    },
+    {
+      "epoch": 94.0074358974359,
+      "grad_norm": 0.0005219983286224306,
+      "learning_rate": 5.84045584045584e-07,
+      "loss": 0.0,
+      "step": 22170
+    },
+    {
+      "epoch": 94.00786324786324,
+      "grad_norm": 0.00013317124103195965,
+      "learning_rate": 5.792972459639127e-07,
+      "loss": 0.0,
+      "step": 22180
+    },
+    {
+      "epoch": 94.0082905982906,
+      "grad_norm": 0.00028317165561020374,
+      "learning_rate": 5.745489078822413e-07,
+      "loss": 0.0,
+      "step": 22190
+    },
+    {
+      "epoch": 94.00871794871794,
+      "grad_norm": 0.00024831280461512506,
+      "learning_rate": 5.698005698005699e-07,
+      "loss": 0.0,
+      "step": 22200
+    },
+    {
+      "epoch": 94.0091452991453,
+      "grad_norm": 24.371335983276367,
+      "learning_rate": 5.650522317188984e-07,
+      "loss": 0.0029,
+      "step": 22210
+    },
+    {
+      "epoch": 94.00957264957265,
+      "grad_norm": 0.0005085449665784836,
+      "learning_rate": 5.60303893637227e-07,
+      "loss": 0.0,
+      "step": 22220
+    },
+    {
+      "epoch": 94.01,
+      "grad_norm": 0.00015020875434856862,
+      "learning_rate": 5.555555555555555e-07,
+      "loss": 0.0,
+      "step": 22230
+    },
+    {
+      "epoch": 94.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.445286273956299,
+      "eval_runtime": 36.5582,
+      "eval_samples_per_second": 0.684,
+      "eval_steps_per_second": 0.684,
+      "step": 22230
+    },
+    {
+      "epoch": 95.00042735042734,
+      "grad_norm": 0.00011259275925112888,
+      "learning_rate": 5.508072174738842e-07,
+      "loss": 0.0,
+      "step": 22240
+    },
+    {
+      "epoch": 95.0008547008547,
+      "grad_norm": 0.00021879606356378645,
+      "learning_rate": 5.460588793922129e-07,
+      "loss": 0.0,
+      "step": 22250
+    },
+    {
+      "epoch": 95.00128205128205,
+      "grad_norm": 0.00042616037535481155,
+      "learning_rate": 5.413105413105414e-07,
+      "loss": 0.0,
+      "step": 22260
+    },
+    {
+      "epoch": 95.0017094017094,
+      "grad_norm": 0.00019269147014711052,
+      "learning_rate": 5.3656220322887e-07,
+      "loss": 0.0,
+      "step": 22270
+    },
+    {
+      "epoch": 95.00213675213675,
+      "grad_norm": 0.00035289855441078544,
+      "learning_rate": 5.318138651471985e-07,
+      "loss": 0.0,
+      "step": 22280
+    },
+    {
+      "epoch": 95.00256410256411,
+      "grad_norm": 0.00012624288501683623,
+      "learning_rate": 5.270655270655271e-07,
+      "loss": 0.0,
+      "step": 22290
+    },
+    {
+      "epoch": 95.00299145299145,
+      "grad_norm": 0.00018576315778773278,
+      "learning_rate": 5.223171889838557e-07,
+      "loss": 0.0,
+      "step": 22300
+    },
+    {
+      "epoch": 95.00341880341881,
+      "grad_norm": 0.000108227992313914,
+      "learning_rate": 5.175688509021843e-07,
+      "loss": 0.0002,
+      "step": 22310
+    },
+    {
+      "epoch": 95.00384615384615,
+      "grad_norm": 0.0001419824839103967,
+      "learning_rate": 5.128205128205128e-07,
+      "loss": 0.0,
+      "step": 22320
+    },
+    {
+      "epoch": 95.0042735042735,
+      "grad_norm": 0.0004882781649939716,
+      "learning_rate": 5.080721747388414e-07,
+      "loss": 0.0,
+      "step": 22330
+    },
+    {
+      "epoch": 95.00470085470086,
+      "grad_norm": 0.0001368028315482661,
+      "learning_rate": 5.0332383665717e-07,
+      "loss": 0.0,
+      "step": 22340
+    },
+    {
+      "epoch": 95.0051282051282,
+      "grad_norm": 0.0009662008378654718,
+      "learning_rate": 4.985754985754987e-07,
+      "loss": 0.0,
+      "step": 22350
+    },
+    {
+      "epoch": 95.00555555555556,
+      "grad_norm": 0.0002917400561273098,
+      "learning_rate": 4.938271604938272e-07,
+      "loss": 0.0,
+      "step": 22360
+    },
+    {
+      "epoch": 95.0059829059829,
+      "grad_norm": 0.00049038470024243,
+      "learning_rate": 4.890788224121558e-07,
+      "loss": 0.0,
+      "step": 22370
+    },
+    {
+      "epoch": 95.00641025641026,
+      "grad_norm": 0.00018329080194234848,
+      "learning_rate": 4.843304843304843e-07,
+      "loss": 0.0,
+      "step": 22380
+    },
+    {
+      "epoch": 95.0068376068376,
+      "grad_norm": 0.00012125585635658354,
+      "learning_rate": 4.795821462488129e-07,
+      "loss": 0.0,
+      "step": 22390
+    },
+    {
+      "epoch": 95.00726495726495,
+      "grad_norm": 0.00016512990987394005,
+      "learning_rate": 4.748338081671416e-07,
+      "loss": 0.0,
+      "step": 22400
+    },
+    {
+      "epoch": 95.00769230769231,
+      "grad_norm": 0.0014239164302125573,
+      "learning_rate": 4.700854700854701e-07,
+      "loss": 0.0,
+      "step": 22410
+    },
+    {
+      "epoch": 95.00811965811965,
+      "grad_norm": 0.011379273608326912,
+      "learning_rate": 4.653371320037987e-07,
+      "loss": 0.0,
+      "step": 22420
+    },
+    {
+      "epoch": 95.00854700854701,
+      "grad_norm": 0.0001255527458852157,
+      "learning_rate": 4.605887939221273e-07,
+      "loss": 0.0,
+      "step": 22430
+    },
+    {
+      "epoch": 95.00897435897436,
+      "grad_norm": 0.00010715150710893795,
+      "learning_rate": 4.558404558404559e-07,
+      "loss": 0.0,
+      "step": 22440
+    },
+    {
+      "epoch": 95.00940170940171,
+      "grad_norm": 0.0006061216117814183,
+      "learning_rate": 4.5109211775878445e-07,
+      "loss": 0.0009,
+      "step": 22450
+    },
+    {
+      "epoch": 95.00982905982906,
+      "grad_norm": 0.0009303994593210518,
+      "learning_rate": 4.4634377967711306e-07,
+      "loss": 0.0,
+      "step": 22460
+    },
+    {
+      "epoch": 95.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.910160064697266,
+      "eval_runtime": 36.7216,
+      "eval_samples_per_second": 0.681,
+      "eval_steps_per_second": 0.681,
+      "step": 22464
+    },
+    {
+      "epoch": 96.00025641025641,
+      "grad_norm": 0.00010953476885333657,
+      "learning_rate": 4.415954415954416e-07,
+      "loss": 0.0,
+      "step": 22470
+    },
+    {
+      "epoch": 96.00068376068376,
+      "grad_norm": 0.0005640748422592878,
+      "learning_rate": 4.3684710351377024e-07,
+      "loss": 0.0,
+      "step": 22480
+    },
+    {
+      "epoch": 96.00111111111111,
+      "grad_norm": 0.00019859550229739398,
+      "learning_rate": 4.320987654320988e-07,
+      "loss": 0.0,
+      "step": 22490
+    },
+    {
+      "epoch": 96.00153846153846,
+      "grad_norm": 0.00011656359856715426,
+      "learning_rate": 4.273504273504274e-07,
+      "loss": 0.0,
+      "step": 22500
+    },
+    {
+      "epoch": 96.00196581196582,
+      "grad_norm": 0.010922577232122421,
+      "learning_rate": 4.226020892687559e-07,
+      "loss": 0.0,
+      "step": 22510
+    },
+    {
+      "epoch": 96.00239316239316,
+      "grad_norm": 0.0017669210210442543,
+      "learning_rate": 4.1785375118708454e-07,
+      "loss": 0.0,
+      "step": 22520
+    },
+    {
+      "epoch": 96.0028205128205,
+      "grad_norm": 9.618465264793485e-05,
+      "learning_rate": 4.131054131054131e-07,
+      "loss": 0.0,
+      "step": 22530
+    },
+    {
+      "epoch": 96.00324786324786,
+      "grad_norm": 0.00018471429939381778,
+      "learning_rate": 4.083570750237417e-07,
+      "loss": 0.0,
+      "step": 22540
+    },
+    {
+      "epoch": 96.00367521367521,
+      "grad_norm": 0.00048411861644126475,
+      "learning_rate": 4.0360873694207033e-07,
+      "loss": 0.0,
+      "step": 22550
+    },
+    {
+      "epoch": 96.00410256410257,
+      "grad_norm": 0.00014513205678667873,
+      "learning_rate": 3.988603988603989e-07,
+      "loss": 0.0,
+      "step": 22560
+    },
+    {
+      "epoch": 96.00452991452991,
+      "grad_norm": 0.00017483255942352116,
+      "learning_rate": 3.941120607787275e-07,
+      "loss": 0.0,
+      "step": 22570
+    },
+    {
+      "epoch": 96.00495726495727,
+      "grad_norm": 0.0007405714131891727,
+      "learning_rate": 3.8936372269705607e-07,
+      "loss": 0.0,
+      "step": 22580
+    },
+    {
+      "epoch": 96.00538461538461,
+      "grad_norm": 0.00010351594391977414,
+      "learning_rate": 3.846153846153847e-07,
+      "loss": 0.0,
+      "step": 22590
+    },
+    {
+      "epoch": 96.00581196581197,
+      "grad_norm": 0.00017762374773155898,
+      "learning_rate": 3.7986704653371324e-07,
+      "loss": 0.0,
+      "step": 22600
+    },
+    {
+      "epoch": 96.00623931623932,
+      "grad_norm": 0.0008599003194831312,
+      "learning_rate": 3.7511870845204186e-07,
+      "loss": 0.0,
+      "step": 22610
+    },
+    {
+      "epoch": 96.00666666666666,
+      "grad_norm": 9.716143540572375e-05,
+      "learning_rate": 3.7037037037037036e-07,
+      "loss": 0.0,
+      "step": 22620
+    },
+    {
+      "epoch": 96.00709401709402,
+      "grad_norm": 0.0006491367239505053,
+      "learning_rate": 3.65622032288699e-07,
+      "loss": 0.0001,
+      "step": 22630
+    },
+    {
+      "epoch": 96.00752136752136,
+      "grad_norm": 0.0002162454038625583,
+      "learning_rate": 3.6087369420702754e-07,
+      "loss": 0.0,
+      "step": 22640
+    },
+    {
+      "epoch": 96.00794871794872,
+      "grad_norm": 0.0001346368808299303,
+      "learning_rate": 3.5612535612535615e-07,
+      "loss": 0.0,
+      "step": 22650
+    },
+    {
+      "epoch": 96.00837606837607,
+      "grad_norm": 0.000659442157484591,
+      "learning_rate": 3.513770180436847e-07,
+      "loss": 0.0,
+      "step": 22660
+    },
+    {
+      "epoch": 96.00880341880342,
+      "grad_norm": 0.00012783669808413833,
+      "learning_rate": 3.4662867996201333e-07,
+      "loss": 0.0,
+      "step": 22670
+    },
+    {
+      "epoch": 96.00923076923077,
+      "grad_norm": 0.00012355089711491019,
+      "learning_rate": 3.4188034188034194e-07,
+      "loss": 0.0,
+      "step": 22680
+    },
+    {
+      "epoch": 96.00965811965811,
+      "grad_norm": 0.00010718983685364947,
+      "learning_rate": 3.371320037986705e-07,
+      "loss": 0.0,
+      "step": 22690
+    },
+    {
+      "epoch": 96.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.9261651039123535,
+      "eval_runtime": 37.5649,
+      "eval_samples_per_second": 0.666,
+      "eval_steps_per_second": 0.666,
+      "step": 22698
+    },
+    {
+      "epoch": 97.00008547008547,
+      "grad_norm": 0.00010189504973823205,
+      "learning_rate": 3.323836657169991e-07,
+      "loss": 0.0,
+      "step": 22700
+    },
+    {
+      "epoch": 97.00051282051282,
+      "grad_norm": 0.01328178122639656,
+      "learning_rate": 3.276353276353277e-07,
+      "loss": 0.4979,
+      "step": 22710
+    },
+    {
+      "epoch": 97.00094017094017,
+      "grad_norm": 0.00010843879863386974,
+      "learning_rate": 3.228869895536563e-07,
+      "loss": 0.0,
+      "step": 22720
+    },
+    {
+      "epoch": 97.00136752136753,
+      "grad_norm": 0.0005735277663916349,
+      "learning_rate": 3.181386514719848e-07,
+      "loss": 0.0,
+      "step": 22730
+    },
+    {
+      "epoch": 97.00179487179487,
+      "grad_norm": 0.0010003969073295593,
+      "learning_rate": 3.133903133903134e-07,
+      "loss": 0.0,
+      "step": 22740
+    },
+    {
+      "epoch": 97.00222222222222,
+      "grad_norm": 0.0009289424051530659,
+      "learning_rate": 3.08641975308642e-07,
+      "loss": 0.0,
+      "step": 22750
+    },
+    {
+      "epoch": 97.00264957264957,
+      "grad_norm": 0.0013247294118627906,
+      "learning_rate": 3.038936372269706e-07,
+      "loss": 0.0,
+      "step": 22760
+    },
+    {
+      "epoch": 97.00307692307692,
+      "grad_norm": 0.00010199108510278165,
+      "learning_rate": 2.991452991452992e-07,
+      "loss": 0.0,
+      "step": 22770
+    },
+    {
+      "epoch": 97.00350427350428,
+      "grad_norm": 0.0014735024888068438,
+      "learning_rate": 2.9439696106362777e-07,
+      "loss": 0.0,
+      "step": 22780
+    },
+    {
+      "epoch": 97.00393162393162,
+      "grad_norm": 0.00014969809853937477,
+      "learning_rate": 2.8964862298195633e-07,
+      "loss": 0.0,
+      "step": 22790
+    },
+    {
+      "epoch": 97.00435897435898,
+      "grad_norm": 0.0001152329205069691,
+      "learning_rate": 2.8490028490028494e-07,
+      "loss": 0.0,
+      "step": 22800
+    },
+    {
+      "epoch": 97.00478632478632,
+      "grad_norm": 0.0011336614843457937,
+      "learning_rate": 2.801519468186135e-07,
+      "loss": 0.0,
+      "step": 22810
+    },
+    {
+      "epoch": 97.00521367521368,
+      "grad_norm": 0.0002444600104354322,
+      "learning_rate": 2.754036087369421e-07,
+      "loss": 0.0,
+      "step": 22820
+    },
+    {
+      "epoch": 97.00564102564103,
+      "grad_norm": 0.0011971363564953208,
+      "learning_rate": 2.706552706552707e-07,
+      "loss": 0.0,
+      "step": 22830
+    },
+    {
+      "epoch": 97.00606837606837,
+      "grad_norm": 0.000934019626583904,
+      "learning_rate": 2.6590693257359924e-07,
+      "loss": 0.0,
+      "step": 22840
+    },
+    {
+      "epoch": 97.00649572649573,
+      "grad_norm": 0.00019381985475774854,
+      "learning_rate": 2.6115859449192786e-07,
+      "loss": 0.0,
+      "step": 22850
+    },
+    {
+      "epoch": 97.00692307692307,
+      "grad_norm": 0.00011170632933499292,
+      "learning_rate": 2.564102564102564e-07,
+      "loss": 0.0,
+      "step": 22860
+    },
+    {
+      "epoch": 97.00735042735043,
+      "grad_norm": 0.00010790720989461988,
+      "learning_rate": 2.51661918328585e-07,
+      "loss": 0.0,
+      "step": 22870
+    },
+    {
+      "epoch": 97.00777777777778,
+      "grad_norm": 0.0001576395152369514,
+      "learning_rate": 2.469135802469136e-07,
+      "loss": 0.0,
+      "step": 22880
+    },
+    {
+      "epoch": 97.00820512820513,
+      "grad_norm": 0.00010023313370766118,
+      "learning_rate": 2.4216524216524215e-07,
+      "loss": 0.0,
+      "step": 22890
+    },
+    {
+      "epoch": 97.00863247863248,
+      "grad_norm": 0.0016645913710817695,
+      "learning_rate": 2.374169040835708e-07,
+      "loss": 0.0,
+      "step": 22900
+    },
+    {
+      "epoch": 97.00905982905982,
+      "grad_norm": 0.0006122203776612878,
+      "learning_rate": 2.3266856600189936e-07,
+      "loss": 0.0,
+      "step": 22910
+    },
+    {
+      "epoch": 97.00948717948718,
+      "grad_norm": 0.0002639706071931869,
+      "learning_rate": 2.2792022792022794e-07,
+      "loss": 0.0,
+      "step": 22920
+    },
+    {
+      "epoch": 97.00991452991452,
+      "grad_norm": 0.00010460082557983696,
+      "learning_rate": 2.2317188983855653e-07,
+      "loss": 0.0,
+      "step": 22930
+    },
+    {
+      "epoch": 97.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.775699615478516,
+      "eval_runtime": 36.4004,
+      "eval_samples_per_second": 0.687,
+      "eval_steps_per_second": 0.687,
+      "step": 22932
+    },
+    {
+      "epoch": 98.00034188034188,
+      "grad_norm": 0.0007676715613342822,
+      "learning_rate": 2.1842355175688512e-07,
+      "loss": 0.0,
+      "step": 22940
+    },
+    {
+      "epoch": 98.00076923076924,
+      "grad_norm": 0.00042605900671333075,
+      "learning_rate": 2.136752136752137e-07,
+      "loss": 0.0,
+      "step": 22950
+    },
+    {
+      "epoch": 98.00119658119658,
+      "grad_norm": 0.0001429203402949497,
+      "learning_rate": 2.0892687559354227e-07,
+      "loss": 0.0,
+      "step": 22960
+    },
+    {
+      "epoch": 98.00162393162393,
+      "grad_norm": 0.0004025712551083416,
+      "learning_rate": 2.0417853751187086e-07,
+      "loss": 0.0,
+      "step": 22970
+    },
+    {
+      "epoch": 98.00205128205128,
+      "grad_norm": 0.001206466811709106,
+      "learning_rate": 1.9943019943019944e-07,
+      "loss": 0.0,
+      "step": 22980
+    },
+    {
+      "epoch": 98.00247863247863,
+      "grad_norm": 0.00012541662727016956,
+      "learning_rate": 1.9468186134852803e-07,
+      "loss": 0.3638,
+      "step": 22990
+    },
+    {
+      "epoch": 98.00290598290599,
+      "grad_norm": 0.0002750447019934654,
+      "learning_rate": 1.8993352326685662e-07,
+      "loss": 0.0,
+      "step": 23000
+    },
+    {
+      "epoch": 98.00333333333333,
+      "grad_norm": 0.00010290797217749059,
+      "learning_rate": 1.8518518518518518e-07,
+      "loss": 0.0,
+      "step": 23010
+    },
+    {
+      "epoch": 98.00376068376069,
+      "grad_norm": 0.0001165928115369752,
+      "learning_rate": 1.8043684710351377e-07,
+      "loss": 0.0,
+      "step": 23020
+    },
+    {
+      "epoch": 98.00418803418803,
+      "grad_norm": 0.0001025166129693389,
+      "learning_rate": 1.7568850902184236e-07,
+      "loss": 0.0,
+      "step": 23030
+    },
+    {
+      "epoch": 98.00461538461539,
+      "grad_norm": 9.691870218375698e-05,
+      "learning_rate": 1.7094017094017097e-07,
+      "loss": 0.0,
+      "step": 23040
+    },
+    {
+      "epoch": 98.00504273504274,
+      "grad_norm": 0.00016511735157109797,
+      "learning_rate": 1.6619183285849956e-07,
+      "loss": 0.0,
+      "step": 23050
+    },
+    {
+      "epoch": 98.00547008547008,
+      "grad_norm": 0.00010559390648268163,
+      "learning_rate": 1.6144349477682815e-07,
+      "loss": 0.0003,
+      "step": 23060
+    },
+    {
+      "epoch": 98.00589743589744,
+      "grad_norm": 0.0008746449020691216,
+      "learning_rate": 1.566951566951567e-07,
+      "loss": 0.0,
+      "step": 23070
+    },
+    {
+      "epoch": 98.00632478632478,
+      "grad_norm": 0.0002250542602268979,
+      "learning_rate": 1.519468186134853e-07,
+      "loss": 0.0,
+      "step": 23080
+    },
+    {
+      "epoch": 98.00675213675214,
+      "grad_norm": 0.00017459446098655462,
+      "learning_rate": 1.4719848053181388e-07,
+      "loss": 0.0,
+      "step": 23090
+    },
+    {
+      "epoch": 98.00717948717949,
+      "grad_norm": 0.00018693608581088483,
+      "learning_rate": 1.4245014245014247e-07,
+      "loss": 0.0,
+      "step": 23100
+    },
+    {
+      "epoch": 98.00760683760684,
+      "grad_norm": 0.0016856353031471372,
+      "learning_rate": 1.3770180436847106e-07,
+      "loss": 0.0,
+      "step": 23110
+    },
+    {
+      "epoch": 98.00803418803419,
+      "grad_norm": 0.00015905691543594003,
+      "learning_rate": 1.3295346628679962e-07,
+      "loss": 0.0,
+      "step": 23120
+    },
+    {
+      "epoch": 98.00846153846153,
+      "grad_norm": 0.00016817098367027938,
+      "learning_rate": 1.282051282051282e-07,
+      "loss": 0.0,
+      "step": 23130
+    },
+    {
+      "epoch": 98.00888888888889,
+      "grad_norm": 0.00019543507369235158,
+      "learning_rate": 1.234567901234568e-07,
+      "loss": 0.0,
+      "step": 23140
+    },
+    {
+      "epoch": 98.00931623931623,
+      "grad_norm": 0.0006627991679124534,
+      "learning_rate": 1.187084520417854e-07,
+      "loss": 0.0,
+      "step": 23150
+    },
+    {
+      "epoch": 98.0097435897436,
+      "grad_norm": 0.00031626957934349775,
+      "learning_rate": 1.1396011396011397e-07,
+      "loss": 0.0,
+      "step": 23160
+    },
+    {
+      "epoch": 98.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.829818248748779,
+      "eval_runtime": 35.756,
+      "eval_samples_per_second": 0.699,
+      "eval_steps_per_second": 0.699,
+      "step": 23166
+    },
+    {
+      "epoch": 99.00017094017095,
+      "grad_norm": 0.00048812164459377527,
+      "learning_rate": 1.0921177587844256e-07,
+      "loss": 0.0,
+      "step": 23170
+    },
+    {
+      "epoch": 99.00059829059829,
+      "grad_norm": 0.0003830118221230805,
+      "learning_rate": 1.0446343779677113e-07,
+      "loss": 0.0,
+      "step": 23180
+    },
+    {
+      "epoch": 99.00102564102563,
+      "grad_norm": 0.00040612902375869453,
+      "learning_rate": 9.971509971509972e-08,
+      "loss": 0.0,
+      "step": 23190
+    },
+    {
+      "epoch": 99.001452991453,
+      "grad_norm": 0.00012840931594837457,
+      "learning_rate": 9.496676163342831e-08,
+      "loss": 0.0,
+      "step": 23200
+    },
+    {
+      "epoch": 99.00188034188034,
+      "grad_norm": 0.0002664999628905207,
+      "learning_rate": 9.021842355175688e-08,
+      "loss": 0.0,
+      "step": 23210
+    },
+    {
+      "epoch": 99.0023076923077,
+      "grad_norm": 0.00010337825369788334,
+      "learning_rate": 8.547008547008549e-08,
+      "loss": 0.0,
+      "step": 23220
+    },
+    {
+      "epoch": 99.00273504273504,
+      "grad_norm": 0.0008002725662663579,
+      "learning_rate": 8.072174738841407e-08,
+      "loss": 0.0,
+      "step": 23230
+    },
+    {
+      "epoch": 99.0031623931624,
+      "grad_norm": 0.00012994400458410382,
+      "learning_rate": 7.597340930674265e-08,
+      "loss": 0.0,
+      "step": 23240
+    },
+    {
+      "epoch": 99.00358974358974,
+      "grad_norm": 0.0011528537143021822,
+      "learning_rate": 7.122507122507124e-08,
+      "loss": 0.0,
+      "step": 23250
+    },
+    {
+      "epoch": 99.0040170940171,
+      "grad_norm": 0.00013322003360372037,
+      "learning_rate": 6.647673314339981e-08,
+      "loss": 0.0,
+      "step": 23260
+    },
+    {
+      "epoch": 99.00444444444445,
+      "grad_norm": 0.001754231285303831,
+      "learning_rate": 6.17283950617284e-08,
+      "loss": 0.0,
+      "step": 23270
+    },
+    {
+      "epoch": 99.00487179487179,
+      "grad_norm": 0.0002544873859733343,
+      "learning_rate": 5.6980056980056986e-08,
+      "loss": 0.0,
+      "step": 23280
+    },
+    {
+      "epoch": 99.00529914529915,
+      "grad_norm": 0.0005114562809467316,
+      "learning_rate": 5.223171889838557e-08,
+      "loss": 0.0,
+      "step": 23290
+    },
+    {
+      "epoch": 99.00572649572649,
+      "grad_norm": 0.00014945660950616002,
+      "learning_rate": 4.7483380816714155e-08,
+      "loss": 0.0,
+      "step": 23300
+    },
+    {
+      "epoch": 99.00615384615385,
+      "grad_norm": 0.0005400913069024682,
+      "learning_rate": 4.273504273504274e-08,
+      "loss": 0.0,
+      "step": 23310
+    },
+    {
+      "epoch": 99.0065811965812,
+      "grad_norm": 0.0001350509119220078,
+      "learning_rate": 3.7986704653371324e-08,
+      "loss": 0.0,
+      "step": 23320
+    },
+    {
+      "epoch": 99.00700854700855,
+      "grad_norm": 0.00013444873911794275,
+      "learning_rate": 3.3238366571699905e-08,
+      "loss": 0.0,
+      "step": 23330
+    },
+    {
+      "epoch": 99.0074358974359,
+      "grad_norm": 0.0003640491340775043,
+      "learning_rate": 2.8490028490028493e-08,
+      "loss": 0.0,
+      "step": 23340
+    },
+    {
+      "epoch": 99.00786324786324,
+      "grad_norm": 0.0004577758372761309,
+      "learning_rate": 2.3741690408357078e-08,
+      "loss": 0.0,
+      "step": 23350
+    },
+    {
+      "epoch": 99.0082905982906,
+      "grad_norm": 0.0005503903958015144,
+      "learning_rate": 1.8993352326685662e-08,
+      "loss": 0.0,
+      "step": 23360
+    },
+    {
+      "epoch": 99.00871794871794,
+      "grad_norm": 0.0008972916402854025,
+      "learning_rate": 1.4245014245014247e-08,
+      "loss": 0.0,
+      "step": 23370
+    },
+    {
+      "epoch": 99.0091452991453,
+      "grad_norm": 0.00013151808525435627,
+      "learning_rate": 9.496676163342831e-09,
+      "loss": 0.0,
+      "step": 23380
+    },
+    {
+      "epoch": 99.00957264957265,
+      "grad_norm": 0.00012301822425797582,
+      "learning_rate": 4.7483380816714155e-09,
+      "loss": 0.0,
+      "step": 23390
+    },
+    {
+      "epoch": 99.01,
+      "grad_norm": 0.002225738950073719,
+      "learning_rate": 0.0,
+      "loss": 0.0,
+      "step": 23400
+    },
+    {
+      "epoch": 99.01,
+      "eval_accuracy": 0.4,
+      "eval_loss": 6.831691741943359,
+      "eval_runtime": 36.0994,
+      "eval_samples_per_second": 0.693,
+      "eval_steps_per_second": 0.693,
+      "step": 23400
+    },
+    {
+      "epoch": 99.01,
+      "step": 23400,
+      "total_flos": 1.0275070897744773e+20,
+      "train_loss": 0.44932733817713766,
+      "train_runtime": 74953.482,
+      "train_samples_per_second": 0.312,
+      "train_steps_per_second": 0.312
+    },
+    {
+      "epoch": 99.01,
+      "eval_accuracy": 0.48,
+      "eval_loss": 4.4508891105651855,
+      "eval_runtime": 32.6922,
+      "eval_samples_per_second": 0.765,
+      "eval_steps_per_second": 0.765,
+      "step": 23400
+    },
+    {
+      "epoch": 99.01,
+      "eval_accuracy": 0.48,
+      "eval_loss": 4.450888633728027,
+      "eval_runtime": 32.6242,
+      "eval_samples_per_second": 0.766,
+      "eval_steps_per_second": 0.766,
+      "step": 23400
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 23400,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.0275070897744773e+20,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}